[PATCH v3 00/16] Move vcpu_load and vcpu_put calls to arch code

[PATCH v3 00/16] Move vcpu_load and vcpu_put calls to arch code

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Some architectures may decide to do different things during
kvm_arch_vcpu_load depending on the ioctl being executed.  For example,
arm64 is about to do significant work in vcpu load/put when running a
vcpu, but it's problematic to do this for any other vcpu ioctl than
KVM_RUN.

Further, while it may be possible to call kvm_arch_vcpu_load() for a
number of non-KVM_RUN ioctls, it makes the KVM/ARM code more difficult
to reason about, especially after my optimization series, because a lot
of things can now happen, where we have to consider if we're really in
the process of running a vcpu or not.

This series will first move the vcpu_load() and vcpu_put() calls in the
arch generic dispatch function into each case of the switch statement
and then, one-by-one, pushed the calls down into the architecture
specific code making the changes for each ioctl as required.

Based on v4.15-rc1

Patches also available at:
git://git.kernel.org/pub/scm/linux/kernel/git/cdall/linux.git vcpu-load-put-v3

Changes since v2:
 - Clarified commit message on patch 1
 - Initialized ret to -EINVAL at declaration on patch 9
 - Added David Hildenbrand's reviewed-by tag

Changes since v1:
 - Fix PPC and S390 bugs from v1
 - Take the mutex in the main disaptcher function and make vcpu_load a
   void, which simplifies the patches overall.
 - Add a patch that moves vcpu_load for arm/arm64 after the first-run
   init function.

Thanks,
-Christoffer

Christoffer Dall (16):
  KVM: Take vcpu->mutex outside vcpu_load
  KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate
  KVM: Move vcpu_load to arch-specific
    kvm_arch_vcpu_ioctl_set_guest_debug
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl
  KVM: arm/arm64: Avoid vcpu_load for other vcpu ioctls than KVM_RUN
  KVM: arm/arm64: Move vcpu_load call after kvm_vcpu_first_run_init

 arch/arm64/kvm/guest.c        |  12 +++--
 arch/mips/kvm/mips.c          |  58 +++++++++++++++--------
 arch/powerpc/kvm/book3s.c     |  24 +++++++++-
 arch/powerpc/kvm/booke.c      |  51 +++++++++++++++-----
 arch/powerpc/kvm/powerpc.c    |  19 +++++---
 arch/s390/kvm/kvm-s390.c      |  90 +++++++++++++++++++++++++++--------
 arch/x86/kvm/vmx.c            |   4 +-
 arch/x86/kvm/x86.c            | 106 ++++++++++++++++++++++++++++++------------
 include/linux/kvm_host.h      |   2 +-
 virt/kvm/arm/arch_timer.c     |   4 --
 virt/kvm/arm/arm.c            |  68 ++++++++++++++++++---------
 virt/kvm/arm/vgic/vgic-init.c |  11 -----
 virt/kvm/kvm_main.c           |  17 ++-----
 13 files changed, 322 insertions(+), 144 deletions(-)

-- 
2.14.2

[PATCH v3 00/16] Move vcpu_load and vcpu_put calls to arch code

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Some architectures may decide to do different things during
kvm_arch_vcpu_load depending on the ioctl being executed.  For example,
arm64 is about to do significant work in vcpu load/put when running a
vcpu, but it's problematic to do this for any other vcpu ioctl than
KVM_RUN.

Further, while it may be possible to call kvm_arch_vcpu_load() for a
number of non-KVM_RUN ioctls, it makes the KVM/ARM code more difficult
to reason about, especially after my optimization series, because a lot
of things can now happen, where we have to consider if we're really in
the process of running a vcpu or not.

This series will first move the vcpu_load() and vcpu_put() calls in the
arch generic dispatch function into each case of the switch statement
and then, one-by-one, pushed the calls down into the architecture
specific code making the changes for each ioctl as required.

Based on v4.15-rc1

Patches also available at:
git://git.kernel.org/pub/scm/linux/kernel/git/cdall/linux.git vcpu-load-put-v3

Changes since v2:
 - Clarified commit message on patch 1
 - Initialized ret to -EINVAL at declaration on patch 9
 - Added David Hildenbrand's reviewed-by tag

Changes since v1:
 - Fix PPC and S390 bugs from v1
 - Take the mutex in the main disaptcher function and make vcpu_load a
   void, which simplifies the patches overall.
 - Add a patch that moves vcpu_load for arm/arm64 after the first-run
   init function.

Thanks,
-Christoffer

Christoffer Dall (16):
  KVM: Take vcpu->mutex outside vcpu_load
  KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate
  KVM: Move vcpu_load to arch-specific
    kvm_arch_vcpu_ioctl_set_guest_debug
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl
  KVM: arm/arm64: Avoid vcpu_load for other vcpu ioctls than KVM_RUN
  KVM: arm/arm64: Move vcpu_load call after kvm_vcpu_first_run_init

 arch/arm64/kvm/guest.c        |  12 +++--
 arch/mips/kvm/mips.c          |  58 +++++++++++++++--------
 arch/powerpc/kvm/book3s.c     |  24 +++++++++-
 arch/powerpc/kvm/booke.c      |  51 +++++++++++++++-----
 arch/powerpc/kvm/powerpc.c    |  19 +++++---
 arch/s390/kvm/kvm-s390.c      |  90 +++++++++++++++++++++++++++--------
 arch/x86/kvm/vmx.c            |   4 +-
 arch/x86/kvm/x86.c            | 106 ++++++++++++++++++++++++++++++------------
 include/linux/kvm_host.h      |   2 +-
 virt/kvm/arm/arch_timer.c     |   4 --
 virt/kvm/arm/arm.c            |  68 ++++++++++++++++++---------
 virt/kvm/arm/vgic/vgic-init.c |  11 -----
 virt/kvm/kvm_main.c           |  17 ++-----
 13 files changed, 322 insertions(+), 144 deletions(-)

-- 
2.14.2

[PATCH v3 00/16] Move vcpu_load and vcpu_put calls to arch code

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Some architectures may decide to do different things during
kvm_arch_vcpu_load depending on the ioctl being executed.  For example,
arm64 is about to do significant work in vcpu load/put when running a
vcpu, but it's problematic to do this for any other vcpu ioctl than
KVM_RUN.

Further, while it may be possible to call kvm_arch_vcpu_load() for a
number of non-KVM_RUN ioctls, it makes the KVM/ARM code more difficult
to reason about, especially after my optimization series, because a lot
of things can now happen, where we have to consider if we're really in
the process of running a vcpu or not.

This series will first move the vcpu_load() and vcpu_put() calls in the
arch generic dispatch function into each case of the switch statement
and then, one-by-one, pushed the calls down into the architecture
specific code making the changes for each ioctl as required.

Based on v4.15-rc1

Patches also available at:
git://git.kernel.org/pub/scm/linux/kernel/git/cdall/linux.git vcpu-load-put-v3

Changes since v2:
 - Clarified commit message on patch 1
 - Initialized ret to -EINVAL at declaration on patch 9
 - Added David Hildenbrand's reviewed-by tag

Changes since v1:
 - Fix PPC and S390 bugs from v1
 - Take the mutex in the main disaptcher function and make vcpu_load a
   void, which simplifies the patches overall.
 - Add a patch that moves vcpu_load for arm/arm64 after the first-run
   init function.

Thanks,
-Christoffer

Christoffer Dall (16):
  KVM: Take vcpu->mutex outside vcpu_load
  KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate
  KVM: Move vcpu_load to arch-specific
    kvm_arch_vcpu_ioctl_set_guest_debug
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl
  KVM: arm/arm64: Avoid vcpu_load for other vcpu ioctls than KVM_RUN
  KVM: arm/arm64: Move vcpu_load call after kvm_vcpu_first_run_init

 arch/arm64/kvm/guest.c        |  12 +++--
 arch/mips/kvm/mips.c          |  58 +++++++++++++++--------
 arch/powerpc/kvm/book3s.c     |  24 +++++++++-
 arch/powerpc/kvm/booke.c      |  51 +++++++++++++++-----
 arch/powerpc/kvm/powerpc.c    |  19 +++++---
 arch/s390/kvm/kvm-s390.c      |  90 +++++++++++++++++++++++++++--------
 arch/x86/kvm/vmx.c            |   4 +-
 arch/x86/kvm/x86.c            | 106 ++++++++++++++++++++++++++++++------------
 include/linux/kvm_host.h      |   2 +-
 virt/kvm/arm/arch_timer.c     |   4 --
 virt/kvm/arm/arm.c            |  68 ++++++++++++++++++---------
 virt/kvm/arm/vgic/vgic-init.c |  11 -----
 virt/kvm/kvm_main.c           |  17 ++-----
 13 files changed, 322 insertions(+), 144 deletions(-)

-- 
2.14.2

[PATCH v3 00/16] Move vcpu_load and vcpu_put calls to arch code

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Some architectures may decide to do different things during
kvm_arch_vcpu_load depending on the ioctl being executed.  For example,
arm64 is about to do significant work in vcpu load/put when running a
vcpu, but it's problematic to do this for any other vcpu ioctl than
KVM_RUN.

Further, while it may be possible to call kvm_arch_vcpu_load() for a
number of non-KVM_RUN ioctls, it makes the KVM/ARM code more difficult
to reason about, especially after my optimization series, because a lot
of things can now happen, where we have to consider if we're really in
the process of running a vcpu or not.

This series will first move the vcpu_load() and vcpu_put() calls in the
arch generic dispatch function into each case of the switch statement
and then, one-by-one, pushed the calls down into the architecture
specific code making the changes for each ioctl as required.

Based on v4.15-rc1

Patches also available at:
git://git.kernel.org/pub/scm/linux/kernel/git/cdall/linux.git vcpu-load-put-v3

Changes since v2:
 - Clarified commit message on patch 1
 - Initialized ret to -EINVAL at declaration on patch 9
 - Added David Hildenbrand's reviewed-by tag

Changes since v1:
 - Fix PPC and S390 bugs from v1
 - Take the mutex in the main disaptcher function and make vcpu_load a
   void, which simplifies the patches overall.
 - Add a patch that moves vcpu_load for arm/arm64 after the first-run
   init function.

Thanks,
-Christoffer

Christoffer Dall (16):
  KVM: Take vcpu->mutex outside vcpu_load
  KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate
  KVM: Move vcpu_load to arch-specific
    kvm_arch_vcpu_ioctl_set_guest_debug
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu
  KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl
  KVM: arm/arm64: Avoid vcpu_load for other vcpu ioctls than KVM_RUN
  KVM: arm/arm64: Move vcpu_load call after kvm_vcpu_first_run_init

 arch/arm64/kvm/guest.c        |  12 +++--
 arch/mips/kvm/mips.c          |  58 +++++++++++++++--------
 arch/powerpc/kvm/book3s.c     |  24 +++++++++-
 arch/powerpc/kvm/booke.c      |  51 +++++++++++++++-----
 arch/powerpc/kvm/powerpc.c    |  19 +++++---
 arch/s390/kvm/kvm-s390.c      |  90 +++++++++++++++++++++++++++--------
 arch/x86/kvm/vmx.c            |   4 +-
 arch/x86/kvm/x86.c            | 106 ++++++++++++++++++++++++++++++------------
 include/linux/kvm_host.h      |   2 +-
 virt/kvm/arm/arch_timer.c     |   4 --
 virt/kvm/arm/arm.c            |  68 ++++++++++++++++++---------
 virt/kvm/arm/vgic/vgic-init.c |  11 -----
 virt/kvm/kvm_main.c           |  17 ++-----
 13 files changed, 322 insertions(+), 144 deletions(-)

-- 
2.14.2

[PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

As we're about to call vcpu_load() from architecture-specific
implementations of the KVM vcpu ioctls, but yet we access data
structures protected by the vcpu->mutex in the generic code, factor
this logic out from vcpu_load().

x86 is the only architecture which calls vcpu_load() outside of the main
vcpu ioctl function, and these calls will no longer take the vcpu mutex
following this patch.  However, with the exception of
kvm_arch_vcpu_postcreate (see below), the callers are either in the
creation or destruction path of the VCPU, which means there cannot be
any concurrent access to the data structure, because the file descriptor
is not yet accessible, or is already gone.

kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
accessible by other in-kernel threads through the kvm->vcpus array, and
we therefore take the vcpu mutex in this case directly.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/x86/kvm/vmx.c       |  4 +---
 arch/x86/kvm/x86.c       | 20 +++++++-------------
 include/linux/kvm_host.h |  2 +-
 virt/kvm/kvm_main.c      | 17 ++++++-----------
 4 files changed, 15 insertions(+), 28 deletions(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 714a0673ec3c..e7c46d20e186 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -9559,10 +9559,8 @@ static void vmx_switch_vmcs(struct kvm_vcpu *vcpu, struct loaded_vmcs *vmcs)
 static void vmx_free_vcpu_nested(struct kvm_vcpu *vcpu)
 {
        struct vcpu_vmx *vmx = to_vmx(vcpu);
-       int r;
 
-       r = vcpu_load(vcpu);
-       BUG_ON(r);
+       vcpu_load(vcpu);
        vmx_switch_vmcs(vcpu, &vmx->vmcs01);
        free_nested(vmx);
        vcpu_put(vcpu);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 34c85aa2e2d1..9b8f864243c9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7747,16 +7747,12 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
 
 int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 {
-	int r;
-
 	kvm_vcpu_mtrr_init(vcpu);
-	r = vcpu_load(vcpu);
-	if (r)
-		return r;
+	vcpu_load(vcpu);
 	kvm_vcpu_reset(vcpu, false);
 	kvm_mmu_setup(vcpu);
 	vcpu_put(vcpu);
-	return r;
+	return 0;
 }
 
 void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
@@ -7766,13 +7762,15 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 	kvm_hv_vcpu_postcreate(vcpu);
 
-	if (vcpu_load(vcpu))
+	if (mutex_lock_killable(&vcpu->mutex))
 		return;
+	vcpu_load(vcpu);
 	msr.data = 0x0;
 	msr.index = MSR_IA32_TSC;
 	msr.host_initiated = true;
 	kvm_write_tsc(vcpu, &msr);
 	vcpu_put(vcpu);
+	mutex_unlock(&vcpu->mutex);
 
 	if (!kvmclock_periodic_sync)
 		return;
@@ -7783,11 +7781,9 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 {
-	int r;
 	vcpu->arch.apf.msr_val = 0;
 
-	r = vcpu_load(vcpu);
-	BUG_ON(r);
+	vcpu_load(vcpu);
 	kvm_mmu_unload(vcpu);
 	vcpu_put(vcpu);
 
@@ -8155,9 +8151,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 
 static void kvm_unload_vcpu_mmu(struct kvm_vcpu *vcpu)
 {
-	int r;
-	r = vcpu_load(vcpu);
-	BUG_ON(r);
+	vcpu_load(vcpu);
 	kvm_mmu_unload(vcpu);
 	vcpu_put(vcpu);
 }
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 2e754b7c282c..a000dd8b75f0 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -533,7 +533,7 @@ static inline int kvm_vcpu_get_idx(struct kvm_vcpu *vcpu)
 int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
 void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);
 
-int __must_check vcpu_load(struct kvm_vcpu *vcpu);
+void vcpu_load(struct kvm_vcpu *vcpu);
 void vcpu_put(struct kvm_vcpu *vcpu);
 
 #ifdef __KVM_HAVE_IOAPIC
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f169ecc4f2e8..39961fb8aef7 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -146,17 +146,12 @@ bool kvm_is_reserved_pfn(kvm_pfn_t pfn)
 /*
  * Switches to specified vcpu, until a matching vcpu_put()
  */
-int vcpu_load(struct kvm_vcpu *vcpu)
+void vcpu_load(struct kvm_vcpu *vcpu)
 {
-	int cpu;
-
-	if (mutex_lock_killable(&vcpu->mutex))
-		return -EINTR;
-	cpu = get_cpu();
+	int cpu = get_cpu();
 	preempt_notifier_register(&vcpu->preempt_notifier);
 	kvm_arch_vcpu_load(vcpu, cpu);
 	put_cpu();
-	return 0;
 }
 EXPORT_SYMBOL_GPL(vcpu_load);
 
@@ -166,7 +161,6 @@ void vcpu_put(struct kvm_vcpu *vcpu)
 	kvm_arch_vcpu_put(vcpu);
 	preempt_notifier_unregister(&vcpu->preempt_notifier);
 	preempt_enable();
-	mutex_unlock(&vcpu->mutex);
 }
 EXPORT_SYMBOL_GPL(vcpu_put);
 
@@ -2529,9 +2523,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 #endif
 
 
-	r = vcpu_load(vcpu);
-	if (r)
-		return r;
+	if (mutex_lock_killable(&vcpu->mutex))
+		return -EINTR;
+	vcpu_load(vcpu);
 	switch (ioctl) {
 	case KVM_RUN: {
 		struct pid *oldpid;
@@ -2704,6 +2698,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 	}
 out:
 	vcpu_put(vcpu);
+	mutex_unlock(&vcpu->mutex);
 	kfree(fpu);
 	kfree(kvm_sregs);
 	return r;
-- 
2.14.2

[PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

As we're about to call vcpu_load() from architecture-specific
implementations of the KVM vcpu ioctls, but yet we access data
structures protected by the vcpu->mutex in the generic code, factor
this logic out from vcpu_load().

x86 is the only architecture which calls vcpu_load() outside of the main
vcpu ioctl function, and these calls will no longer take the vcpu mutex
following this patch.  However, with the exception of
kvm_arch_vcpu_postcreate (see below), the callers are either in the
creation or destruction path of the VCPU, which means there cannot be
any concurrent access to the data structure, because the file descriptor
is not yet accessible, or is already gone.

kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
accessible by other in-kernel threads through the kvm->vcpus array, and
we therefore take the vcpu mutex in this case directly.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/x86/kvm/vmx.c       |  4 +---
 arch/x86/kvm/x86.c       | 20 +++++++-------------
 include/linux/kvm_host.h |  2 +-
 virt/kvm/kvm_main.c      | 17 ++++++-----------
 4 files changed, 15 insertions(+), 28 deletions(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 714a0673ec3c..e7c46d20e186 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -9559,10 +9559,8 @@ static void vmx_switch_vmcs(struct kvm_vcpu *vcpu, struct loaded_vmcs *vmcs)
 static void vmx_free_vcpu_nested(struct kvm_vcpu *vcpu)
 {
        struct vcpu_vmx *vmx = to_vmx(vcpu);
-       int r;
 
-       r = vcpu_load(vcpu);
-       BUG_ON(r);
+       vcpu_load(vcpu);
        vmx_switch_vmcs(vcpu, &vmx->vmcs01);
        free_nested(vmx);
        vcpu_put(vcpu);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 34c85aa2e2d1..9b8f864243c9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7747,16 +7747,12 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
 
 int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 {
-	int r;
-
 	kvm_vcpu_mtrr_init(vcpu);
-	r = vcpu_load(vcpu);
-	if (r)
-		return r;
+	vcpu_load(vcpu);
 	kvm_vcpu_reset(vcpu, false);
 	kvm_mmu_setup(vcpu);
 	vcpu_put(vcpu);
-	return r;
+	return 0;
 }
 
 void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
@@ -7766,13 +7762,15 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 	kvm_hv_vcpu_postcreate(vcpu);
 
-	if (vcpu_load(vcpu))
+	if (mutex_lock_killable(&vcpu->mutex))
 		return;
+	vcpu_load(vcpu);
 	msr.data = 0x0;
 	msr.index = MSR_IA32_TSC;
 	msr.host_initiated = true;
 	kvm_write_tsc(vcpu, &msr);
 	vcpu_put(vcpu);
+	mutex_unlock(&vcpu->mutex);
 
 	if (!kvmclock_periodic_sync)
 		return;
@@ -7783,11 +7781,9 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 {
-	int r;
 	vcpu->arch.apf.msr_val = 0;
 
-	r = vcpu_load(vcpu);
-	BUG_ON(r);
+	vcpu_load(vcpu);
 	kvm_mmu_unload(vcpu);
 	vcpu_put(vcpu);
 
@@ -8155,9 +8151,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 
 static void kvm_unload_vcpu_mmu(struct kvm_vcpu *vcpu)
 {
-	int r;
-	r = vcpu_load(vcpu);
-	BUG_ON(r);
+	vcpu_load(vcpu);
 	kvm_mmu_unload(vcpu);
 	vcpu_put(vcpu);
 }
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 2e754b7c282c..a000dd8b75f0 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -533,7 +533,7 @@ static inline int kvm_vcpu_get_idx(struct kvm_vcpu *vcpu)
 int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
 void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);
 
-int __must_check vcpu_load(struct kvm_vcpu *vcpu);
+void vcpu_load(struct kvm_vcpu *vcpu);
 void vcpu_put(struct kvm_vcpu *vcpu);
 
 #ifdef __KVM_HAVE_IOAPIC
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f169ecc4f2e8..39961fb8aef7 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -146,17 +146,12 @@ bool kvm_is_reserved_pfn(kvm_pfn_t pfn)
 /*
  * Switches to specified vcpu, until a matching vcpu_put()
  */
-int vcpu_load(struct kvm_vcpu *vcpu)
+void vcpu_load(struct kvm_vcpu *vcpu)
 {
-	int cpu;
-
-	if (mutex_lock_killable(&vcpu->mutex))
-		return -EINTR;
-	cpu = get_cpu();
+	int cpu = get_cpu();
 	preempt_notifier_register(&vcpu->preempt_notifier);
 	kvm_arch_vcpu_load(vcpu, cpu);
 	put_cpu();
-	return 0;
 }
 EXPORT_SYMBOL_GPL(vcpu_load);
 
@@ -166,7 +161,6 @@ void vcpu_put(struct kvm_vcpu *vcpu)
 	kvm_arch_vcpu_put(vcpu);
 	preempt_notifier_unregister(&vcpu->preempt_notifier);
 	preempt_enable();
-	mutex_unlock(&vcpu->mutex);
 }
 EXPORT_SYMBOL_GPL(vcpu_put);
 
@@ -2529,9 +2523,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 #endif
 
 
-	r = vcpu_load(vcpu);
-	if (r)
-		return r;
+	if (mutex_lock_killable(&vcpu->mutex))
+		return -EINTR;
+	vcpu_load(vcpu);
 	switch (ioctl) {
 	case KVM_RUN: {
 		struct pid *oldpid;
@@ -2704,6 +2698,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 	}
 out:
 	vcpu_put(vcpu);
+	mutex_unlock(&vcpu->mutex);
 	kfree(fpu);
 	kfree(kvm_sregs);
 	return r;
-- 
2.14.2

[PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

As we're about to call vcpu_load() from architecture-specific
implementations of the KVM vcpu ioctls, but yet we access data
structures protected by the vcpu->mutex in the generic code, factor
this logic out from vcpu_load().

x86 is the only architecture which calls vcpu_load() outside of the main
vcpu ioctl function, and these calls will no longer take the vcpu mutex
following this patch.  However, with the exception of
kvm_arch_vcpu_postcreate (see below), the callers are either in the
creation or destruction path of the VCPU, which means there cannot be
any concurrent access to the data structure, because the file descriptor
is not yet accessible, or is already gone.

kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
accessible by other in-kernel threads through the kvm->vcpus array, and
we therefore take the vcpu mutex in this case directly.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/x86/kvm/vmx.c       |  4 +---
 arch/x86/kvm/x86.c       | 20 +++++++-------------
 include/linux/kvm_host.h |  2 +-
 virt/kvm/kvm_main.c      | 17 ++++++-----------
 4 files changed, 15 insertions(+), 28 deletions(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 714a0673ec3c..e7c46d20e186 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -9559,10 +9559,8 @@ static void vmx_switch_vmcs(struct kvm_vcpu *vcpu, struct loaded_vmcs *vmcs)
 static void vmx_free_vcpu_nested(struct kvm_vcpu *vcpu)
 {
        struct vcpu_vmx *vmx = to_vmx(vcpu);
-       int r;
 
-       r = vcpu_load(vcpu);
-       BUG_ON(r);
+       vcpu_load(vcpu);
        vmx_switch_vmcs(vcpu, &vmx->vmcs01);
        free_nested(vmx);
        vcpu_put(vcpu);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 34c85aa2e2d1..9b8f864243c9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7747,16 +7747,12 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
 
 int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 {
-	int r;
-
 	kvm_vcpu_mtrr_init(vcpu);
-	r = vcpu_load(vcpu);
-	if (r)
-		return r;
+	vcpu_load(vcpu);
 	kvm_vcpu_reset(vcpu, false);
 	kvm_mmu_setup(vcpu);
 	vcpu_put(vcpu);
-	return r;
+	return 0;
 }
 
 void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
@@ -7766,13 +7762,15 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 	kvm_hv_vcpu_postcreate(vcpu);
 
-	if (vcpu_load(vcpu))
+	if (mutex_lock_killable(&vcpu->mutex))
 		return;
+	vcpu_load(vcpu);
 	msr.data = 0x0;
 	msr.index = MSR_IA32_TSC;
 	msr.host_initiated = true;
 	kvm_write_tsc(vcpu, &msr);
 	vcpu_put(vcpu);
+	mutex_unlock(&vcpu->mutex);
 
 	if (!kvmclock_periodic_sync)
 		return;
@@ -7783,11 +7781,9 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 {
-	int r;
 	vcpu->arch.apf.msr_val = 0;
 
-	r = vcpu_load(vcpu);
-	BUG_ON(r);
+	vcpu_load(vcpu);
 	kvm_mmu_unload(vcpu);
 	vcpu_put(vcpu);
 
@@ -8155,9 +8151,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 
 static void kvm_unload_vcpu_mmu(struct kvm_vcpu *vcpu)
 {
-	int r;
-	r = vcpu_load(vcpu);
-	BUG_ON(r);
+	vcpu_load(vcpu);
 	kvm_mmu_unload(vcpu);
 	vcpu_put(vcpu);
 }
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 2e754b7c282c..a000dd8b75f0 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -533,7 +533,7 @@ static inline int kvm_vcpu_get_idx(struct kvm_vcpu *vcpu)
 int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
 void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);
 
-int __must_check vcpu_load(struct kvm_vcpu *vcpu);
+void vcpu_load(struct kvm_vcpu *vcpu);
 void vcpu_put(struct kvm_vcpu *vcpu);
 
 #ifdef __KVM_HAVE_IOAPIC
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f169ecc4f2e8..39961fb8aef7 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -146,17 +146,12 @@ bool kvm_is_reserved_pfn(kvm_pfn_t pfn)
 /*
  * Switches to specified vcpu, until a matching vcpu_put()
  */
-int vcpu_load(struct kvm_vcpu *vcpu)
+void vcpu_load(struct kvm_vcpu *vcpu)
 {
-	int cpu;
-
-	if (mutex_lock_killable(&vcpu->mutex))
-		return -EINTR;
-	cpu = get_cpu();
+	int cpu = get_cpu();
 	preempt_notifier_register(&vcpu->preempt_notifier);
 	kvm_arch_vcpu_load(vcpu, cpu);
 	put_cpu();
-	return 0;
 }
 EXPORT_SYMBOL_GPL(vcpu_load);
 
@@ -166,7 +161,6 @@ void vcpu_put(struct kvm_vcpu *vcpu)
 	kvm_arch_vcpu_put(vcpu);
 	preempt_notifier_unregister(&vcpu->preempt_notifier);
 	preempt_enable();
-	mutex_unlock(&vcpu->mutex);
 }
 EXPORT_SYMBOL_GPL(vcpu_put);
 
@@ -2529,9 +2523,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 #endif
 
 
-	r = vcpu_load(vcpu);
-	if (r)
-		return r;
+	if (mutex_lock_killable(&vcpu->mutex))
+		return -EINTR;
+	vcpu_load(vcpu);
 	switch (ioctl) {
 	case KVM_RUN: {
 		struct pid *oldpid;
@@ -2704,6 +2698,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 	}
 out:
 	vcpu_put(vcpu);
+	mutex_unlock(&vcpu->mutex);
 	kfree(fpu);
 	kfree(kvm_sregs);
 	return r;
-- 
2.14.2

[PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

As we're about to call vcpu_load() from architecture-specific
implementations of the KVM vcpu ioctls, but yet we access data
structures protected by the vcpu->mutex in the generic code, factor
this logic out from vcpu_load().

x86 is the only architecture which calls vcpu_load() outside of the main
vcpu ioctl function, and these calls will no longer take the vcpu mutex
following this patch.  However, with the exception of
kvm_arch_vcpu_postcreate (see below), the callers are either in the
creation or destruction path of the VCPU, which means there cannot be
any concurrent access to the data structure, because the file descriptor
is not yet accessible, or is already gone.

kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
accessible by other in-kernel threads through the kvm->vcpus array, and
we therefore take the vcpu mutex in this case directly.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/x86/kvm/vmx.c       |  4 +---
 arch/x86/kvm/x86.c       | 20 +++++++-------------
 include/linux/kvm_host.h |  2 +-
 virt/kvm/kvm_main.c      | 17 ++++++-----------
 4 files changed, 15 insertions(+), 28 deletions(-)

diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 714a0673ec3c..e7c46d20e186 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -9559,10 +9559,8 @@ static void vmx_switch_vmcs(struct kvm_vcpu *vcpu, struct loaded_vmcs *vmcs)
 static void vmx_free_vcpu_nested(struct kvm_vcpu *vcpu)
 {
        struct vcpu_vmx *vmx = to_vmx(vcpu);
-       int r;
 
-       r = vcpu_load(vcpu);
-       BUG_ON(r);
+       vcpu_load(vcpu);
        vmx_switch_vmcs(vcpu, &vmx->vmcs01);
        free_nested(vmx);
        vcpu_put(vcpu);
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 34c85aa2e2d1..9b8f864243c9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7747,16 +7747,12 @@ struct kvm_vcpu *kvm_arch_vcpu_create(struct kvm *kvm,
 
 int kvm_arch_vcpu_setup(struct kvm_vcpu *vcpu)
 {
-	int r;
-
 	kvm_vcpu_mtrr_init(vcpu);
-	r = vcpu_load(vcpu);
-	if (r)
-		return r;
+	vcpu_load(vcpu);
 	kvm_vcpu_reset(vcpu, false);
 	kvm_mmu_setup(vcpu);
 	vcpu_put(vcpu);
-	return r;
+	return 0;
 }
 
 void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
@@ -7766,13 +7762,15 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 	kvm_hv_vcpu_postcreate(vcpu);
 
-	if (vcpu_load(vcpu))
+	if (mutex_lock_killable(&vcpu->mutex))
 		return;
+	vcpu_load(vcpu);
 	msr.data = 0x0;
 	msr.index = MSR_IA32_TSC;
 	msr.host_initiated = true;
 	kvm_write_tsc(vcpu, &msr);
 	vcpu_put(vcpu);
+	mutex_unlock(&vcpu->mutex);
 
 	if (!kvmclock_periodic_sync)
 		return;
@@ -7783,11 +7781,9 @@ void kvm_arch_vcpu_postcreate(struct kvm_vcpu *vcpu)
 
 void kvm_arch_vcpu_destroy(struct kvm_vcpu *vcpu)
 {
-	int r;
 	vcpu->arch.apf.msr_val = 0;
 
-	r = vcpu_load(vcpu);
-	BUG_ON(r);
+	vcpu_load(vcpu);
 	kvm_mmu_unload(vcpu);
 	vcpu_put(vcpu);
 
@@ -8155,9 +8151,7 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 
 static void kvm_unload_vcpu_mmu(struct kvm_vcpu *vcpu)
 {
-	int r;
-	r = vcpu_load(vcpu);
-	BUG_ON(r);
+	vcpu_load(vcpu);
 	kvm_mmu_unload(vcpu);
 	vcpu_put(vcpu);
 }
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 2e754b7c282c..a000dd8b75f0 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -533,7 +533,7 @@ static inline int kvm_vcpu_get_idx(struct kvm_vcpu *vcpu)
 int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id);
 void kvm_vcpu_uninit(struct kvm_vcpu *vcpu);
 
-int __must_check vcpu_load(struct kvm_vcpu *vcpu);
+void vcpu_load(struct kvm_vcpu *vcpu);
 void vcpu_put(struct kvm_vcpu *vcpu);
 
 #ifdef __KVM_HAVE_IOAPIC
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f169ecc4f2e8..39961fb8aef7 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -146,17 +146,12 @@ bool kvm_is_reserved_pfn(kvm_pfn_t pfn)
 /*
  * Switches to specified vcpu, until a matching vcpu_put()
  */
-int vcpu_load(struct kvm_vcpu *vcpu)
+void vcpu_load(struct kvm_vcpu *vcpu)
 {
-	int cpu;
-
-	if (mutex_lock_killable(&vcpu->mutex))
-		return -EINTR;
-	cpu = get_cpu();
+	int cpu = get_cpu();
 	preempt_notifier_register(&vcpu->preempt_notifier);
 	kvm_arch_vcpu_load(vcpu, cpu);
 	put_cpu();
-	return 0;
 }
 EXPORT_SYMBOL_GPL(vcpu_load);
 
@@ -166,7 +161,6 @@ void vcpu_put(struct kvm_vcpu *vcpu)
 	kvm_arch_vcpu_put(vcpu);
 	preempt_notifier_unregister(&vcpu->preempt_notifier);
 	preempt_enable();
-	mutex_unlock(&vcpu->mutex);
 }
 EXPORT_SYMBOL_GPL(vcpu_put);
 
@@ -2529,9 +2523,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 #endif
 
 
-	r = vcpu_load(vcpu);
-	if (r)
-		return r;
+	if (mutex_lock_killable(&vcpu->mutex))
+		return -EINTR;
+	vcpu_load(vcpu);
 	switch (ioctl) {
 	case KVM_RUN: {
 		struct pid *oldpid;
@@ -2704,6 +2698,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 	}
 out:
 	vcpu_put(vcpu);
+	mutex_unlock(&vcpu->mutex);
 	kfree(fpu);
 	kfree(kvm_sregs);
 	return r;
-- 
2.14.2

Re: [PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Christian Borntraeger]

On 12/04/2017 09:35 PM, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> As we're about to call vcpu_load() from architecture-specific
> implementations of the KVM vcpu ioctls, but yet we access data
> structures protected by the vcpu->mutex in the generic code, factor
> this logic out from vcpu_load().
> 
> x86 is the only architecture which calls vcpu_load() outside of the main
> vcpu ioctl function, and these calls will no longer take the vcpu mutex
> following this patch.  However, with the exception of
> kvm_arch_vcpu_postcreate (see below), the callers are either in the
> creation or destruction path of the VCPU, which means there cannot be
> any concurrent access to the data structure, because the file descriptor
> is not yet accessible, or is already gone.
> 
> kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
> accessible by other in-kernel threads through the kvm->vcpus array, and
> we therefore take the vcpu mutex in this case directly.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>

Looks good to me.

Re: [PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Christian Borntraeger]

On 12/04/2017 09:35 PM, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> As we're about to call vcpu_load() from architecture-specific
> implementations of the KVM vcpu ioctls, but yet we access data
> structures protected by the vcpu->mutex in the generic code, factor
> this logic out from vcpu_load().
> 
> x86 is the only architecture which calls vcpu_load() outside of the main
> vcpu ioctl function, and these calls will no longer take the vcpu mutex
> following this patch.  However, with the exception of
> kvm_arch_vcpu_postcreate (see below), the callers are either in the
> creation or destruction path of the VCPU, which means there cannot be
> any concurrent access to the data structure, because the file descriptor
> is not yet accessible, or is already gone.
> 
> kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
> accessible by other in-kernel threads through the kvm->vcpus array, and
> we therefore take the vcpu mutex in this case directly.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>

Looks good to me.

[PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Christian Borntraeger]

On 12/04/2017 09:35 PM, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> As we're about to call vcpu_load() from architecture-specific
> implementations of the KVM vcpu ioctls, but yet we access data
> structures protected by the vcpu->mutex in the generic code, factor
> this logic out from vcpu_load().
> 
> x86 is the only architecture which calls vcpu_load() outside of the main
> vcpu ioctl function, and these calls will no longer take the vcpu mutex
> following this patch.  However, with the exception of
> kvm_arch_vcpu_postcreate (see below), the callers are either in the
> creation or destruction path of the VCPU, which means there cannot be
> any concurrent access to the data structure, because the file descriptor
> is not yet accessible, or is already gone.
> 
> kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
> accessible by other in-kernel threads through the kvm->vcpus array, and
> we therefore take the vcpu mutex in this case directly.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>

Looks good to me.

Re: [PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:23 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> As we're about to call vcpu_load() from architecture-specific
> implementations of the KVM vcpu ioctls, but yet we access data
> structures protected by the vcpu->mutex in the generic code, factor
> this logic out from vcpu_load().
> 
> x86 is the only architecture which calls vcpu_load() outside of the main
> vcpu ioctl function, and these calls will no longer take the vcpu mutex
> following this patch.  However, with the exception of
> kvm_arch_vcpu_postcreate (see below), the callers are either in the
> creation or destruction path of the VCPU, which means there cannot be
> any concurrent access to the data structure, because the file descriptor
> is not yet accessible, or is already gone.
> 
> kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
> accessible by other in-kernel threads through the kvm->vcpus array, and
> we therefore take the vcpu mutex in this case directly.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/x86/kvm/vmx.c       |  4 +---
>  arch/x86/kvm/x86.c       | 20 +++++++-------------
>  include/linux/kvm_host.h |  2 +-
>  virt/kvm/kvm_main.c      | 17 ++++++-----------
>  4 files changed, 15 insertions(+), 28 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:23 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> As we're about to call vcpu_load() from architecture-specific
> implementations of the KVM vcpu ioctls, but yet we access data
> structures protected by the vcpu->mutex in the generic code, factor
> this logic out from vcpu_load().
> 
> x86 is the only architecture which calls vcpu_load() outside of the main
> vcpu ioctl function, and these calls will no longer take the vcpu mutex
> following this patch.  However, with the exception of
> kvm_arch_vcpu_postcreate (see below), the callers are either in the
> creation or destruction path of the VCPU, which means there cannot be
> any concurrent access to the data structure, because the file descriptor
> is not yet accessible, or is already gone.
> 
> kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
> accessible by other in-kernel threads through the kvm->vcpus array, and
> we therefore take the vcpu mutex in this case directly.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/x86/kvm/vmx.c       |  4 +---
>  arch/x86/kvm/x86.c       | 20 +++++++-------------
>  include/linux/kvm_host.h |  2 +-
>  virt/kvm/kvm_main.c      | 17 ++++++-----------
>  4 files changed, 15 insertions(+), 28 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 01/16] KVM: Take vcpu->mutex outside vcpu_load

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:23 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> As we're about to call vcpu_load() from architecture-specific
> implementations of the KVM vcpu ioctls, but yet we access data
> structures protected by the vcpu->mutex in the generic code, factor
> this logic out from vcpu_load().
> 
> x86 is the only architecture which calls vcpu_load() outside of the main
> vcpu ioctl function, and these calls will no longer take the vcpu mutex
> following this patch.  However, with the exception of
> kvm_arch_vcpu_postcreate (see below), the callers are either in the
> creation or destruction path of the VCPU, which means there cannot be
> any concurrent access to the data structure, because the file descriptor
> is not yet accessible, or is already gone.
> 
> kvm_arch_vcpu_postcreate makes the newly created vcpu potentially
> accessible by other in-kernel threads through the kvm->vcpus array, and
> we therefore take the vcpu mutex in this case directly.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/x86/kvm/vmx.c       |  4 +---
>  arch/x86/kvm/x86.c       | 20 +++++++-------------
>  include/linux/kvm_host.h |  2 +-
>  virt/kvm/kvm_main.c      | 17 ++++++-----------
>  4 files changed, 15 insertions(+), 28 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 02/16] KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

In preparation for moving calls to vcpu_load() and vcpu_put() into the
architecture specific implementations of the KVM vcpu ioctls, move the
calls in the main kvm_vcpu_ioctl() dispatcher function to each case
of the ioctl select statement.  This allows us to move the vcpu_load()
and vcpu_put() calls into architecture specific implementations of vcpu
ioctls, one by one.

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 virt/kvm/kvm_main.c | 26 ++++++++++++++++++++++++--
 1 file changed, 24 insertions(+), 2 deletions(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 39961fb8aef7..480b16c48f6b 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2525,13 +2525,13 @@ static long kvm_vcpu_ioctl(struct file *filp,
 
 	if (mutex_lock_killable(&vcpu->mutex))
 		return -EINTR;
-	vcpu_load(vcpu);
 	switch (ioctl) {
 	case KVM_RUN: {
 		struct pid *oldpid;
 		r = -EINVAL;
 		if (arg)
 			goto out;
+		vcpu_load(vcpu);
 		oldpid = rcu_access_pointer(vcpu->pid);
 		if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
 			/* The thread running this VCPU changed. */
@@ -2543,6 +2543,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			put_pid(oldpid);
 		}
 		r = kvm_arch_vcpu_ioctl_run(vcpu, vcpu->run);
+		vcpu_put(vcpu);
 		trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
 		break;
 	}
@@ -2553,7 +2554,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
 		if (!kvm_regs)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
+		vcpu_put(vcpu);
 		if (r)
 			goto out_free1;
 		r = -EFAULT;
@@ -2573,7 +2576,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			r = PTR_ERR(kvm_regs);
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
+		vcpu_put(vcpu);
 		kfree(kvm_regs);
 		break;
 	}
@@ -2582,7 +2587,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!kvm_sregs)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2598,13 +2605,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			kvm_sregs = NULL;
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_sregs(vcpu, kvm_sregs);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_GET_MP_STATE: {
 		struct kvm_mp_state mp_state;
 
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_mpstate(vcpu, &mp_state);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2619,7 +2630,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&mp_state, argp, sizeof(mp_state)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_mpstate(vcpu, &mp_state);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_TRANSLATE: {
@@ -2628,7 +2641,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&tr, argp, sizeof(tr)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_translate(vcpu, &tr);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2643,7 +2658,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&dbg, argp, sizeof(dbg)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_guest_debug(vcpu, &dbg);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_SET_SIGNAL_MASK: {
@@ -2674,7 +2691,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!fpu)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_fpu(vcpu, fpu);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2690,14 +2709,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			fpu = NULL;
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
+		vcpu_put(vcpu);
 		break;
 	}
 	default:
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
+		vcpu_put(vcpu);
 	}
 out:
-	vcpu_put(vcpu);
 	mutex_unlock(&vcpu->mutex);
 	kfree(fpu);
 	kfree(kvm_sregs);
-- 
2.14.2

[PATCH v3 02/16] KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

In preparation for moving calls to vcpu_load() and vcpu_put() into the
architecture specific implementations of the KVM vcpu ioctls, move the
calls in the main kvm_vcpu_ioctl() dispatcher function to each case
of the ioctl select statement.  This allows us to move the vcpu_load()
and vcpu_put() calls into architecture specific implementations of vcpu
ioctls, one by one.

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 virt/kvm/kvm_main.c | 26 ++++++++++++++++++++++++--
 1 file changed, 24 insertions(+), 2 deletions(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 39961fb8aef7..480b16c48f6b 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2525,13 +2525,13 @@ static long kvm_vcpu_ioctl(struct file *filp,
 
 	if (mutex_lock_killable(&vcpu->mutex))
 		return -EINTR;
-	vcpu_load(vcpu);
 	switch (ioctl) {
 	case KVM_RUN: {
 		struct pid *oldpid;
 		r = -EINVAL;
 		if (arg)
 			goto out;
+		vcpu_load(vcpu);
 		oldpid = rcu_access_pointer(vcpu->pid);
 		if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
 			/* The thread running this VCPU changed. */
@@ -2543,6 +2543,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			put_pid(oldpid);
 		}
 		r = kvm_arch_vcpu_ioctl_run(vcpu, vcpu->run);
+		vcpu_put(vcpu);
 		trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
 		break;
 	}
@@ -2553,7 +2554,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
 		if (!kvm_regs)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
+		vcpu_put(vcpu);
 		if (r)
 			goto out_free1;
 		r = -EFAULT;
@@ -2573,7 +2576,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			r = PTR_ERR(kvm_regs);
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
+		vcpu_put(vcpu);
 		kfree(kvm_regs);
 		break;
 	}
@@ -2582,7 +2587,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!kvm_sregs)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2598,13 +2605,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			kvm_sregs = NULL;
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_sregs(vcpu, kvm_sregs);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_GET_MP_STATE: {
 		struct kvm_mp_state mp_state;
 
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_mpstate(vcpu, &mp_state);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2619,7 +2630,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&mp_state, argp, sizeof(mp_state)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_mpstate(vcpu, &mp_state);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_TRANSLATE: {
@@ -2628,7 +2641,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&tr, argp, sizeof(tr)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_translate(vcpu, &tr);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2643,7 +2658,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&dbg, argp, sizeof(dbg)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_guest_debug(vcpu, &dbg);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_SET_SIGNAL_MASK: {
@@ -2674,7 +2691,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!fpu)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_fpu(vcpu, fpu);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2690,14 +2709,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			fpu = NULL;
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
+		vcpu_put(vcpu);
 		break;
 	}
 	default:
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
+		vcpu_put(vcpu);
 	}
 out:
-	vcpu_put(vcpu);
 	mutex_unlock(&vcpu->mutex);
 	kfree(fpu);
 	kfree(kvm_sregs);
-- 
2.14.2

[PATCH v3 02/16] KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

In preparation for moving calls to vcpu_load() and vcpu_put() into the
architecture specific implementations of the KVM vcpu ioctls, move the
calls in the main kvm_vcpu_ioctl() dispatcher function to each case
of the ioctl select statement.  This allows us to move the vcpu_load()
and vcpu_put() calls into architecture specific implementations of vcpu
ioctls, one by one.

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 virt/kvm/kvm_main.c | 26 ++++++++++++++++++++++++--
 1 file changed, 24 insertions(+), 2 deletions(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 39961fb8aef7..480b16c48f6b 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2525,13 +2525,13 @@ static long kvm_vcpu_ioctl(struct file *filp,
 
 	if (mutex_lock_killable(&vcpu->mutex))
 		return -EINTR;
-	vcpu_load(vcpu);
 	switch (ioctl) {
 	case KVM_RUN: {
 		struct pid *oldpid;
 		r = -EINVAL;
 		if (arg)
 			goto out;
+		vcpu_load(vcpu);
 		oldpid = rcu_access_pointer(vcpu->pid);
 		if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
 			/* The thread running this VCPU changed. */
@@ -2543,6 +2543,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			put_pid(oldpid);
 		}
 		r = kvm_arch_vcpu_ioctl_run(vcpu, vcpu->run);
+		vcpu_put(vcpu);
 		trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
 		break;
 	}
@@ -2553,7 +2554,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
 		if (!kvm_regs)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
+		vcpu_put(vcpu);
 		if (r)
 			goto out_free1;
 		r = -EFAULT;
@@ -2573,7 +2576,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			r = PTR_ERR(kvm_regs);
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
+		vcpu_put(vcpu);
 		kfree(kvm_regs);
 		break;
 	}
@@ -2582,7 +2587,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!kvm_sregs)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2598,13 +2605,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			kvm_sregs = NULL;
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_sregs(vcpu, kvm_sregs);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_GET_MP_STATE: {
 		struct kvm_mp_state mp_state;
 
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_mpstate(vcpu, &mp_state);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2619,7 +2630,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&mp_state, argp, sizeof(mp_state)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_mpstate(vcpu, &mp_state);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_TRANSLATE: {
@@ -2628,7 +2641,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&tr, argp, sizeof(tr)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_translate(vcpu, &tr);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2643,7 +2658,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&dbg, argp, sizeof(dbg)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_guest_debug(vcpu, &dbg);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_SET_SIGNAL_MASK: {
@@ -2674,7 +2691,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!fpu)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_fpu(vcpu, fpu);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2690,14 +2709,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			fpu = NULL;
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
+		vcpu_put(vcpu);
 		break;
 	}
 	default:
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
+		vcpu_put(vcpu);
 	}
 out:
-	vcpu_put(vcpu);
 	mutex_unlock(&vcpu->mutex);
 	kfree(fpu);
 	kfree(kvm_sregs);
-- 
2.14.2

[PATCH v3 02/16] KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

In preparation for moving calls to vcpu_load() and vcpu_put() into the
architecture specific implementations of the KVM vcpu ioctls, move the
calls in the main kvm_vcpu_ioctl() dispatcher function to each case
of the ioctl select statement.  This allows us to move the vcpu_load()
and vcpu_put() calls into architecture specific implementations of vcpu
ioctls, one by one.

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 virt/kvm/kvm_main.c | 26 ++++++++++++++++++++++++--
 1 file changed, 24 insertions(+), 2 deletions(-)

diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 39961fb8aef7..480b16c48f6b 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2525,13 +2525,13 @@ static long kvm_vcpu_ioctl(struct file *filp,
 
 	if (mutex_lock_killable(&vcpu->mutex))
 		return -EINTR;
-	vcpu_load(vcpu);
 	switch (ioctl) {
 	case KVM_RUN: {
 		struct pid *oldpid;
 		r = -EINVAL;
 		if (arg)
 			goto out;
+		vcpu_load(vcpu);
 		oldpid = rcu_access_pointer(vcpu->pid);
 		if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
 			/* The thread running this VCPU changed. */
@@ -2543,6 +2543,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			put_pid(oldpid);
 		}
 		r = kvm_arch_vcpu_ioctl_run(vcpu, vcpu->run);
+		vcpu_put(vcpu);
 		trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
 		break;
 	}
@@ -2553,7 +2554,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
 		if (!kvm_regs)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
+		vcpu_put(vcpu);
 		if (r)
 			goto out_free1;
 		r = -EFAULT;
@@ -2573,7 +2576,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			r = PTR_ERR(kvm_regs);
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
+		vcpu_put(vcpu);
 		kfree(kvm_regs);
 		break;
 	}
@@ -2582,7 +2587,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!kvm_sregs)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2598,13 +2605,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			kvm_sregs = NULL;
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_sregs(vcpu, kvm_sregs);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_GET_MP_STATE: {
 		struct kvm_mp_state mp_state;
 
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_mpstate(vcpu, &mp_state);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2619,7 +2630,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&mp_state, argp, sizeof(mp_state)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_mpstate(vcpu, &mp_state);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_TRANSLATE: {
@@ -2628,7 +2641,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&tr, argp, sizeof(tr)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_translate(vcpu, &tr);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2643,7 +2658,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&dbg, argp, sizeof(dbg)))
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_guest_debug(vcpu, &dbg);
+		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_SET_SIGNAL_MASK: {
@@ -2674,7 +2691,9 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!fpu)
 			goto out;
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_fpu(vcpu, fpu);
+		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
@@ -2690,14 +2709,17 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			fpu = NULL;
 			goto out;
 		}
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
+		vcpu_put(vcpu);
 		break;
 	}
 	default:
+		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
+		vcpu_put(vcpu);
 	}
 out:
-	vcpu_put(vcpu);
 	mutex_unlock(&vcpu->mutex);
 	kfree(fpu);
 	kfree(kvm_sregs);
-- 
2.14.2

Re: [PATCH v3 02/16] KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:24 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> In preparation for moving calls to vcpu_load() and vcpu_put() into the
> architecture specific implementations of the KVM vcpu ioctls, move the
> calls in the main kvm_vcpu_ioctl() dispatcher function to each case
> of the ioctl select statement.  This allows us to move the vcpu_load()
> and vcpu_put() calls into architecture specific implementations of vcpu
> ioctls, one by one.
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  virt/kvm/kvm_main.c | 26 ++++++++++++++++++++++++--
>  1 file changed, 24 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 02/16] KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:24 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> In preparation for moving calls to vcpu_load() and vcpu_put() into the
> architecture specific implementations of the KVM vcpu ioctls, move the
> calls in the main kvm_vcpu_ioctl() dispatcher function to each case
> of the ioctl select statement.  This allows us to move the vcpu_load()
> and vcpu_put() calls into architecture specific implementations of vcpu
> ioctls, one by one.
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  virt/kvm/kvm_main.c | 26 ++++++++++++++++++++++++--
>  1 file changed, 24 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 02/16] KVM: Prepare for moving vcpu_load/vcpu_put into arch specific code

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:24 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> In preparation for moving calls to vcpu_load() and vcpu_put() into the
> architecture specific implementations of the KVM vcpu ioctls, move the
> calls in the main kvm_vcpu_ioctl() dispatcher function to each case
> of the ioctl select statement.  This allows us to move the vcpu_load()
> and vcpu_put() calls into architecture specific implementations of vcpu
> ioctls, one by one.
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  virt/kvm/kvm_main.c | 26 ++++++++++++++++++++++++--
>  1 file changed, 24 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_run().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c       |  3 +++
 arch/powerpc/kvm/powerpc.c |  6 +++++-
 arch/s390/kvm/kvm-s390.c   | 10 ++++++++--
 arch/x86/kvm/x86.c         |  3 +++
 virt/kvm/arm/arm.c         | 15 +++++++++++----
 virt/kvm/kvm_main.c        |  2 --
 6 files changed, 30 insertions(+), 9 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index d535edc01434..b5c28f0730f8 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -447,6 +447,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r = -EINTR;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
 
@@ -483,6 +485,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index 6b6c53c42ac9..c06bc9552438 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1409,6 +1409,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->mmio_needed) {
 		vcpu->mmio_needed = 0;
 		if (!vcpu->mmio_is_write)
@@ -1423,7 +1425,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 			r = kvmppc_emulate_mmio_vsx_loadstore(vcpu, run);
 			if (r == RESUME_HOST) {
 				vcpu->mmio_needed = 1;
-				return r;
+				goto out;
 			}
 		}
 #endif
@@ -1459,6 +1461,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+out:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 98ad8b9e0360..2b3e874ea76c 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -3377,9 +3377,12 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (kvm_run->immediate_exit)
 		return -EINTR;
 
+	vcpu_load(vcpu);
+
 	if (guestdbg_exit_pending(vcpu)) {
 		kvm_s390_prepare_debug_exit(vcpu);
-		return 0;
+		rc = 0;
+		goto out;
 	}
 
 	if (vcpu->sigset_active)
@@ -3390,7 +3393,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	} else if (is_vcpu_stopped(vcpu)) {
 		pr_err_ratelimited("can't run stopped vcpu %d\n",
 				   vcpu->vcpu_id);
-		return -EINVAL;
+		rc = -EINVAL;
+		goto out;
 	}
 
 	sync_regs(vcpu, kvm_run);
@@ -3421,6 +3425,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
 	vcpu->stat.exit_userspace++;
+out:
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9b8f864243c9..d9deb6222055 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7252,6 +7252,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	fpu__initialize(fpu);
 
 	if (vcpu->sigset_active)
@@ -7301,6 +7303,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a6524ff27de4..1f448b208686 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -620,18 +620,22 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (unlikely(!kvm_vcpu_initialized(vcpu)))
 		return -ENOEXEC;
 
+	vcpu_load(vcpu);
+
 	ret = kvm_vcpu_first_run_init(vcpu);
 	if (ret)
-		return ret;
+		goto out;
 
 	if (run->exit_reason == KVM_EXIT_MMIO) {
 		ret = kvm_handle_mmio_return(vcpu, vcpu->run);
 		if (ret)
-			return ret;
+			goto out;
 	}
 
-	if (run->immediate_exit)
-		return -EINTR;
+	if (run->immediate_exit) {
+		ret = -EINTR;
+		goto out;
+	}
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
@@ -771,6 +775,9 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
+
+out:
+	vcpu_put(vcpu);
 	return ret;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 480b16c48f6b..198f2f9edcaf 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2531,7 +2531,6 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 		if (arg)
 			goto out;
-		vcpu_load(vcpu);
 		oldpid = rcu_access_pointer(vcpu->pid);
 		if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
 			/* The thread running this VCPU changed. */
@@ -2543,7 +2542,6 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			put_pid(oldpid);
 		}
 		r = kvm_arch_vcpu_ioctl_run(vcpu, vcpu->run);
-		vcpu_put(vcpu);
 		trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
 		break;
 	}
-- 
2.14.2

[PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_run().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c       |  3 +++
 arch/powerpc/kvm/powerpc.c |  6 +++++-
 arch/s390/kvm/kvm-s390.c   | 10 ++++++++--
 arch/x86/kvm/x86.c         |  3 +++
 virt/kvm/arm/arm.c         | 15 +++++++++++----
 virt/kvm/kvm_main.c        |  2 --
 6 files changed, 30 insertions(+), 9 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index d535edc01434..b5c28f0730f8 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -447,6 +447,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r = -EINTR;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
 
@@ -483,6 +485,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index 6b6c53c42ac9..c06bc9552438 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1409,6 +1409,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->mmio_needed) {
 		vcpu->mmio_needed = 0;
 		if (!vcpu->mmio_is_write)
@@ -1423,7 +1425,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 			r = kvmppc_emulate_mmio_vsx_loadstore(vcpu, run);
 			if (r = RESUME_HOST) {
 				vcpu->mmio_needed = 1;
-				return r;
+				goto out;
 			}
 		}
 #endif
@@ -1459,6 +1461,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+out:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 98ad8b9e0360..2b3e874ea76c 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -3377,9 +3377,12 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (kvm_run->immediate_exit)
 		return -EINTR;
 
+	vcpu_load(vcpu);
+
 	if (guestdbg_exit_pending(vcpu)) {
 		kvm_s390_prepare_debug_exit(vcpu);
-		return 0;
+		rc = 0;
+		goto out;
 	}
 
 	if (vcpu->sigset_active)
@@ -3390,7 +3393,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	} else if (is_vcpu_stopped(vcpu)) {
 		pr_err_ratelimited("can't run stopped vcpu %d\n",
 				   vcpu->vcpu_id);
-		return -EINVAL;
+		rc = -EINVAL;
+		goto out;
 	}
 
 	sync_regs(vcpu, kvm_run);
@@ -3421,6 +3425,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
 	vcpu->stat.exit_userspace++;
+out:
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9b8f864243c9..d9deb6222055 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7252,6 +7252,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	fpu__initialize(fpu);
 
 	if (vcpu->sigset_active)
@@ -7301,6 +7303,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a6524ff27de4..1f448b208686 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -620,18 +620,22 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (unlikely(!kvm_vcpu_initialized(vcpu)))
 		return -ENOEXEC;
 
+	vcpu_load(vcpu);
+
 	ret = kvm_vcpu_first_run_init(vcpu);
 	if (ret)
-		return ret;
+		goto out;
 
 	if (run->exit_reason = KVM_EXIT_MMIO) {
 		ret = kvm_handle_mmio_return(vcpu, vcpu->run);
 		if (ret)
-			return ret;
+			goto out;
 	}
 
-	if (run->immediate_exit)
-		return -EINTR;
+	if (run->immediate_exit) {
+		ret = -EINTR;
+		goto out;
+	}
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
@@ -771,6 +775,9 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
+
+out:
+	vcpu_put(vcpu);
 	return ret;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 480b16c48f6b..198f2f9edcaf 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2531,7 +2531,6 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 		if (arg)
 			goto out;
-		vcpu_load(vcpu);
 		oldpid = rcu_access_pointer(vcpu->pid);
 		if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
 			/* The thread running this VCPU changed. */
@@ -2543,7 +2542,6 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			put_pid(oldpid);
 		}
 		r = kvm_arch_vcpu_ioctl_run(vcpu, vcpu->run);
-		vcpu_put(vcpu);
 		trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
 		break;
 	}
-- 
2.14.2

[PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_run().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c       |  3 +++
 arch/powerpc/kvm/powerpc.c |  6 +++++-
 arch/s390/kvm/kvm-s390.c   | 10 ++++++++--
 arch/x86/kvm/x86.c         |  3 +++
 virt/kvm/arm/arm.c         | 15 +++++++++++----
 virt/kvm/kvm_main.c        |  2 --
 6 files changed, 30 insertions(+), 9 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index d535edc01434..b5c28f0730f8 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -447,6 +447,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r = -EINTR;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
 
@@ -483,6 +485,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index 6b6c53c42ac9..c06bc9552438 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1409,6 +1409,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->mmio_needed) {
 		vcpu->mmio_needed = 0;
 		if (!vcpu->mmio_is_write)
@@ -1423,7 +1425,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 			r = kvmppc_emulate_mmio_vsx_loadstore(vcpu, run);
 			if (r == RESUME_HOST) {
 				vcpu->mmio_needed = 1;
-				return r;
+				goto out;
 			}
 		}
 #endif
@@ -1459,6 +1461,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+out:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 98ad8b9e0360..2b3e874ea76c 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -3377,9 +3377,12 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (kvm_run->immediate_exit)
 		return -EINTR;
 
+	vcpu_load(vcpu);
+
 	if (guestdbg_exit_pending(vcpu)) {
 		kvm_s390_prepare_debug_exit(vcpu);
-		return 0;
+		rc = 0;
+		goto out;
 	}
 
 	if (vcpu->sigset_active)
@@ -3390,7 +3393,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	} else if (is_vcpu_stopped(vcpu)) {
 		pr_err_ratelimited("can't run stopped vcpu %d\n",
 				   vcpu->vcpu_id);
-		return -EINVAL;
+		rc = -EINVAL;
+		goto out;
 	}
 
 	sync_regs(vcpu, kvm_run);
@@ -3421,6 +3425,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
 	vcpu->stat.exit_userspace++;
+out:
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9b8f864243c9..d9deb6222055 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7252,6 +7252,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	fpu__initialize(fpu);
 
 	if (vcpu->sigset_active)
@@ -7301,6 +7303,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a6524ff27de4..1f448b208686 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -620,18 +620,22 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (unlikely(!kvm_vcpu_initialized(vcpu)))
 		return -ENOEXEC;
 
+	vcpu_load(vcpu);
+
 	ret = kvm_vcpu_first_run_init(vcpu);
 	if (ret)
-		return ret;
+		goto out;
 
 	if (run->exit_reason == KVM_EXIT_MMIO) {
 		ret = kvm_handle_mmio_return(vcpu, vcpu->run);
 		if (ret)
-			return ret;
+			goto out;
 	}
 
-	if (run->immediate_exit)
-		return -EINTR;
+	if (run->immediate_exit) {
+		ret = -EINTR;
+		goto out;
+	}
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
@@ -771,6 +775,9 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
+
+out:
+	vcpu_put(vcpu);
 	return ret;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 480b16c48f6b..198f2f9edcaf 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2531,7 +2531,6 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 		if (arg)
 			goto out;
-		vcpu_load(vcpu);
 		oldpid = rcu_access_pointer(vcpu->pid);
 		if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
 			/* The thread running this VCPU changed. */
@@ -2543,7 +2542,6 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			put_pid(oldpid);
 		}
 		r = kvm_arch_vcpu_ioctl_run(vcpu, vcpu->run);
-		vcpu_put(vcpu);
 		trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
 		break;
 	}
-- 
2.14.2

[PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_run().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c       |  3 +++
 arch/powerpc/kvm/powerpc.c |  6 +++++-
 arch/s390/kvm/kvm-s390.c   | 10 ++++++++--
 arch/x86/kvm/x86.c         |  3 +++
 virt/kvm/arm/arm.c         | 15 +++++++++++----
 virt/kvm/kvm_main.c        |  2 --
 6 files changed, 30 insertions(+), 9 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index d535edc01434..b5c28f0730f8 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -447,6 +447,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r = -EINTR;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
 
@@ -483,6 +485,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index 6b6c53c42ac9..c06bc9552438 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1409,6 +1409,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	if (vcpu->mmio_needed) {
 		vcpu->mmio_needed = 0;
 		if (!vcpu->mmio_is_write)
@@ -1423,7 +1425,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 			r = kvmppc_emulate_mmio_vsx_loadstore(vcpu, run);
 			if (r == RESUME_HOST) {
 				vcpu->mmio_needed = 1;
-				return r;
+				goto out;
 			}
 		}
 #endif
@@ -1459,6 +1461,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+out:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 98ad8b9e0360..2b3e874ea76c 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -3377,9 +3377,12 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (kvm_run->immediate_exit)
 		return -EINTR;
 
+	vcpu_load(vcpu);
+
 	if (guestdbg_exit_pending(vcpu)) {
 		kvm_s390_prepare_debug_exit(vcpu);
-		return 0;
+		rc = 0;
+		goto out;
 	}
 
 	if (vcpu->sigset_active)
@@ -3390,7 +3393,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	} else if (is_vcpu_stopped(vcpu)) {
 		pr_err_ratelimited("can't run stopped vcpu %d\n",
 				   vcpu->vcpu_id);
-		return -EINVAL;
+		rc = -EINVAL;
+		goto out;
 	}
 
 	sync_regs(vcpu, kvm_run);
@@ -3421,6 +3425,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
 	vcpu->stat.exit_userspace++;
+out:
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9b8f864243c9..d9deb6222055 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7252,6 +7252,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	int r;
 	sigset_t sigsaved;
 
+	vcpu_load(vcpu);
+
 	fpu__initialize(fpu);
 
 	if (vcpu->sigset_active)
@@ -7301,6 +7303,7 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a6524ff27de4..1f448b208686 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -620,18 +620,22 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (unlikely(!kvm_vcpu_initialized(vcpu)))
 		return -ENOEXEC;
 
+	vcpu_load(vcpu);
+
 	ret = kvm_vcpu_first_run_init(vcpu);
 	if (ret)
-		return ret;
+		goto out;
 
 	if (run->exit_reason == KVM_EXIT_MMIO) {
 		ret = kvm_handle_mmio_return(vcpu, vcpu->run);
 		if (ret)
-			return ret;
+			goto out;
 	}
 
-	if (run->immediate_exit)
-		return -EINTR;
+	if (run->immediate_exit) {
+		ret = -EINTR;
+		goto out;
+	}
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
@@ -771,6 +775,9 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
+
+out:
+	vcpu_put(vcpu);
 	return ret;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 480b16c48f6b..198f2f9edcaf 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2531,7 +2531,6 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 		if (arg)
 			goto out;
-		vcpu_load(vcpu);
 		oldpid = rcu_access_pointer(vcpu->pid);
 		if (unlikely(oldpid != current->pids[PIDTYPE_PID].pid)) {
 			/* The thread running this VCPU changed. */
@@ -2543,7 +2542,6 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			put_pid(oldpid);
 		}
 		r = kvm_arch_vcpu_ioctl_run(vcpu, vcpu->run);
-		vcpu_put(vcpu);
 		trace_kvm_userspace_exit(vcpu->run->exit_reason, r);
 		break;
 	}
-- 
2.14.2

Re: [PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Christian Borntraeger]

On 12/04/2017 09:35 PM, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_run().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>

Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com> # s390 parts

> ---

Re: [PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Christian Borntraeger]

On 12/04/2017 09:35 PM, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_run().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>

Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com> # s390 parts

> ---

[PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Christian Borntraeger]

On 12/04/2017 09:35 PM, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_run().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>

Reviewed-by: Christian Borntraeger <borntraeger@de.ibm.com> # s390 parts

> ---

Re: [PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:25 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_run().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c       |  3 +++
>  arch/powerpc/kvm/powerpc.c |  6 +++++-
>  arch/s390/kvm/kvm-s390.c   | 10 ++++++++--
>  arch/x86/kvm/x86.c         |  3 +++
>  virt/kvm/arm/arm.c         | 15 +++++++++++----
>  virt/kvm/kvm_main.c        |  2 --
>  6 files changed, 30 insertions(+), 9 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:25 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_run().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c       |  3 +++
>  arch/powerpc/kvm/powerpc.c |  6 +++++-
>  arch/s390/kvm/kvm-s390.c   | 10 ++++++++--
>  arch/x86/kvm/x86.c         |  3 +++
>  virt/kvm/arm/arm.c         | 15 +++++++++++----
>  virt/kvm/kvm_main.c        |  2 --
>  6 files changed, 30 insertions(+), 9 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 03/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_run

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:25 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_run().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c       |  3 +++
>  arch/powerpc/kvm/powerpc.c |  6 +++++-
>  arch/s390/kvm/kvm-s390.c   | 10 ++++++++--
>  arch/x86/kvm/x86.c         |  3 +++
>  virt/kvm/arm/arm.c         | 15 +++++++++++----
>  virt/kvm/kvm_main.c        |  2 --
>  6 files changed, 30 insertions(+), 9 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c      | 3 +++
 arch/powerpc/kvm/book3s.c | 3 +++
 arch/powerpc/kvm/booke.c  | 3 +++
 arch/s390/kvm/kvm-s390.c  | 2 ++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 6 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index b5c28f0730f8..adfca57420d1 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -1165,6 +1165,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	for (i = 0; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
 		regs->gpr[i] = vcpu->arch.gprs[i];
 
@@ -1172,6 +1174,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->lo = vcpu->arch.lo;
 	regs->pc = vcpu->arch.pc;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 72d977e30952..d85bfd733ccd 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -497,6 +497,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	regs->pc = kvmppc_get_pc(vcpu);
 	regs->cr = kvmppc_get_cr(vcpu);
 	regs->ctr = kvmppc_get_ctr(vcpu);
@@ -518,6 +520,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 83b485810aea..e0e4f04c5535 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1431,6 +1431,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	regs->pc = vcpu->arch.pc;
 	regs->cr = kvmppc_get_cr(vcpu);
 	regs->ctr = vcpu->arch.ctr;
@@ -1452,6 +1454,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 2b3e874ea76c..37b7caae2484 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2718,7 +2718,9 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
 	memcpy(&regs->gprs, &vcpu->run->s.regs.gprs, sizeof(regs->gprs));
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index d9deb6222055..597e1f8fc8da 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7309,6 +7309,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
+
 	if (vcpu->arch.emulate_regs_need_sync_to_vcpu) {
 		/*
 		 * We are here if userspace calls get_regs() in the middle of
@@ -7342,6 +7344,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->rip = kvm_rip_read(vcpu);
 	regs->rflags = kvm_get_rflags(vcpu);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 198f2f9edcaf..843d481f58cb 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2552,9 +2552,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
 		if (!kvm_regs)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
-		vcpu_put(vcpu);
 		if (r)
 			goto out_free1;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c      | 3 +++
 arch/powerpc/kvm/book3s.c | 3 +++
 arch/powerpc/kvm/booke.c  | 3 +++
 arch/s390/kvm/kvm-s390.c  | 2 ++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 6 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index b5c28f0730f8..adfca57420d1 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -1165,6 +1165,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	for (i = 0; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
 		regs->gpr[i] = vcpu->arch.gprs[i];
 
@@ -1172,6 +1174,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->lo = vcpu->arch.lo;
 	regs->pc = vcpu->arch.pc;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 72d977e30952..d85bfd733ccd 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -497,6 +497,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	regs->pc = kvmppc_get_pc(vcpu);
 	regs->cr = kvmppc_get_cr(vcpu);
 	regs->ctr = kvmppc_get_ctr(vcpu);
@@ -518,6 +520,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 83b485810aea..e0e4f04c5535 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1431,6 +1431,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	regs->pc = vcpu->arch.pc;
 	regs->cr = kvmppc_get_cr(vcpu);
 	regs->ctr = vcpu->arch.ctr;
@@ -1452,6 +1454,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 2b3e874ea76c..37b7caae2484 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2718,7 +2718,9 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
 	memcpy(&regs->gprs, &vcpu->run->s.regs.gprs, sizeof(regs->gprs));
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index d9deb6222055..597e1f8fc8da 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7309,6 +7309,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
+
 	if (vcpu->arch.emulate_regs_need_sync_to_vcpu) {
 		/*
 		 * We are here if userspace calls get_regs() in the middle of
@@ -7342,6 +7344,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->rip = kvm_rip_read(vcpu);
 	regs->rflags = kvm_get_rflags(vcpu);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 198f2f9edcaf..843d481f58cb 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2552,9 +2552,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
 		if (!kvm_regs)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
-		vcpu_put(vcpu);
 		if (r)
 			goto out_free1;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c      | 3 +++
 arch/powerpc/kvm/book3s.c | 3 +++
 arch/powerpc/kvm/booke.c  | 3 +++
 arch/s390/kvm/kvm-s390.c  | 2 ++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 6 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index b5c28f0730f8..adfca57420d1 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -1165,6 +1165,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	for (i = 0; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
 		regs->gpr[i] = vcpu->arch.gprs[i];
 
@@ -1172,6 +1174,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->lo = vcpu->arch.lo;
 	regs->pc = vcpu->arch.pc;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 72d977e30952..d85bfd733ccd 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -497,6 +497,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	regs->pc = kvmppc_get_pc(vcpu);
 	regs->cr = kvmppc_get_cr(vcpu);
 	regs->ctr = kvmppc_get_ctr(vcpu);
@@ -518,6 +520,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 83b485810aea..e0e4f04c5535 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1431,6 +1431,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	regs->pc = vcpu->arch.pc;
 	regs->cr = kvmppc_get_cr(vcpu);
 	regs->ctr = vcpu->arch.ctr;
@@ -1452,6 +1454,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 2b3e874ea76c..37b7caae2484 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2718,7 +2718,9 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
 	memcpy(&regs->gprs, &vcpu->run->s.regs.gprs, sizeof(regs->gprs));
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index d9deb6222055..597e1f8fc8da 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7309,6 +7309,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
+
 	if (vcpu->arch.emulate_regs_need_sync_to_vcpu) {
 		/*
 		 * We are here if userspace calls get_regs() in the middle of
@@ -7342,6 +7344,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->rip = kvm_rip_read(vcpu);
 	regs->rflags = kvm_get_rflags(vcpu);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 198f2f9edcaf..843d481f58cb 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2552,9 +2552,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
 		if (!kvm_regs)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
-		vcpu_put(vcpu);
 		if (r)
 			goto out_free1;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c      | 3 +++
 arch/powerpc/kvm/book3s.c | 3 +++
 arch/powerpc/kvm/booke.c  | 3 +++
 arch/s390/kvm/kvm-s390.c  | 2 ++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 6 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index b5c28f0730f8..adfca57420d1 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -1165,6 +1165,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	for (i = 0; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
 		regs->gpr[i] = vcpu->arch.gprs[i];
 
@@ -1172,6 +1174,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->lo = vcpu->arch.lo;
 	regs->pc = vcpu->arch.pc;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 72d977e30952..d85bfd733ccd 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -497,6 +497,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	regs->pc = kvmppc_get_pc(vcpu);
 	regs->cr = kvmppc_get_cr(vcpu);
 	regs->ctr = kvmppc_get_ctr(vcpu);
@@ -518,6 +520,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 83b485810aea..e0e4f04c5535 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1431,6 +1431,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	regs->pc = vcpu->arch.pc;
 	regs->cr = kvmppc_get_cr(vcpu);
 	regs->ctr = vcpu->arch.ctr;
@@ -1452,6 +1454,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 2b3e874ea76c..37b7caae2484 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2718,7 +2718,9 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
 	memcpy(&regs->gprs, &vcpu->run->s.regs.gprs, sizeof(regs->gprs));
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index d9deb6222055..597e1f8fc8da 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7309,6 +7309,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
+
 	if (vcpu->arch.emulate_regs_need_sync_to_vcpu) {
 		/*
 		 * We are here if userspace calls get_regs() in the middle of
@@ -7342,6 +7344,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	regs->rip = kvm_rip_read(vcpu);
 	regs->rflags = kvm_get_rflags(vcpu);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 198f2f9edcaf..843d481f58cb 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2552,9 +2552,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
 		if (!kvm_regs)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
-		vcpu_put(vcpu);
 		if (r)
 			goto out_free1;
 		r = -EFAULT;
-- 
2.14.2

Re: [PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index b5c28f0730f8..adfca57420d1 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -1165,6 +1165,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	for (i = 0; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
>  		regs->gpr[i] = vcpu->arch.gprs[i];
>  
> @@ -1172,6 +1174,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	regs->lo = vcpu->arch.lo;
>  	regs->pc = vcpu->arch.pc;
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index 72d977e30952..d85bfd733ccd 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -497,6 +497,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	regs->pc = kvmppc_get_pc(vcpu);
>  	regs->cr = kvmppc_get_cr(vcpu);
>  	regs->ctr = kvmppc_get_ctr(vcpu);
> @@ -518,6 +520,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index 83b485810aea..e0e4f04c5535 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1431,6 +1431,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	regs->pc = vcpu->arch.pc;
>  	regs->cr = kvmppc_get_cr(vcpu);
>  	regs->ctr = vcpu->arch.ctr;
> @@ -1452,6 +1454,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 2b3e874ea76c..37b7caae2484 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2718,7 +2718,9 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
>  	memcpy(&regs->gprs, &vcpu->run->s.regs.gprs, sizeof(regs->gprs));
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index d9deb6222055..597e1f8fc8da 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7309,6 +7309,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
> +
>  	if (vcpu->arch.emulate_regs_need_sync_to_vcpu) {
>  		/*
>  		 * We are here if userspace calls get_regs() in the middle of
> @@ -7342,6 +7344,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	regs->rip = kvm_rip_read(vcpu);
>  	regs->rflags = kvm_get_rflags(vcpu);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 198f2f9edcaf..843d481f58cb 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2552,9 +2552,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
>  		if (!kvm_regs)
>  			goto out;
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
> -		vcpu_put(vcpu);
>  		if (r)
>  			goto out_free1;
>  		r = -EFAULT;
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index b5c28f0730f8..adfca57420d1 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -1165,6 +1165,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	for (i = 0; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
>  		regs->gpr[i] = vcpu->arch.gprs[i];
>  
> @@ -1172,6 +1174,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	regs->lo = vcpu->arch.lo;
>  	regs->pc = vcpu->arch.pc;
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index 72d977e30952..d85bfd733ccd 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -497,6 +497,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	regs->pc = kvmppc_get_pc(vcpu);
>  	regs->cr = kvmppc_get_cr(vcpu);
>  	regs->ctr = kvmppc_get_ctr(vcpu);
> @@ -518,6 +520,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index 83b485810aea..e0e4f04c5535 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1431,6 +1431,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	regs->pc = vcpu->arch.pc;
>  	regs->cr = kvmppc_get_cr(vcpu);
>  	regs->ctr = vcpu->arch.ctr;
> @@ -1452,6 +1454,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 2b3e874ea76c..37b7caae2484 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2718,7 +2718,9 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
>  	memcpy(&regs->gprs, &vcpu->run->s.regs.gprs, sizeof(regs->gprs));
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index d9deb6222055..597e1f8fc8da 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7309,6 +7309,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
> +
>  	if (vcpu->arch.emulate_regs_need_sync_to_vcpu) {
>  		/*
>  		 * We are here if userspace calls get_regs() in the middle of
> @@ -7342,6 +7344,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	regs->rip = kvm_rip_read(vcpu);
>  	regs->rflags = kvm_get_rflags(vcpu);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 198f2f9edcaf..843d481f58cb 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2552,9 +2552,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
>  		if (!kvm_regs)
>  			goto out;
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
> -		vcpu_put(vcpu);
>  		if (r)
>  			goto out_free1;
>  		r = -EFAULT;
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

[PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index b5c28f0730f8..adfca57420d1 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -1165,6 +1165,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	for (i = 0; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
>  		regs->gpr[i] = vcpu->arch.gprs[i];
>  
> @@ -1172,6 +1174,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	regs->lo = vcpu->arch.lo;
>  	regs->pc = vcpu->arch.pc;
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index 72d977e30952..d85bfd733ccd 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -497,6 +497,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	regs->pc = kvmppc_get_pc(vcpu);
>  	regs->cr = kvmppc_get_cr(vcpu);
>  	regs->ctr = kvmppc_get_ctr(vcpu);
> @@ -518,6 +520,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index 83b485810aea..e0e4f04c5535 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1431,6 +1431,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	regs->pc = vcpu->arch.pc;
>  	regs->cr = kvmppc_get_cr(vcpu);
>  	regs->ctr = vcpu->arch.ctr;
> @@ -1452,6 +1454,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		regs->gpr[i] = kvmppc_get_gpr(vcpu, i);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 2b3e874ea76c..37b7caae2484 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2718,7 +2718,9 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
>  	memcpy(&regs->gprs, &vcpu->run->s.regs.gprs, sizeof(regs->gprs));
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index d9deb6222055..597e1f8fc8da 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7309,6 +7309,8 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
> +
>  	if (vcpu->arch.emulate_regs_need_sync_to_vcpu) {
>  		/*
>  		 * We are here if userspace calls get_regs() in the middle of
> @@ -7342,6 +7344,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	regs->rip = kvm_rip_read(vcpu);
>  	regs->rflags = kvm_get_rflags(vcpu);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 198f2f9edcaf..843d481f58cb 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2552,9 +2552,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		kvm_regs = kzalloc(sizeof(struct kvm_regs), GFP_KERNEL);
>  		if (!kvm_regs)
>  			goto out;
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_get_regs(vcpu, kvm_regs);
> -		vcpu_put(vcpu);
>  		if (r)
>  			goto out_free1;
>  		r = -EFAULT;
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:26 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:26 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 04/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_regs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:26 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c      | 3 +++
 arch/powerpc/kvm/book3s.c | 3 +++
 arch/powerpc/kvm/booke.c  | 3 +++
 arch/s390/kvm/kvm-s390.c  | 2 ++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 6 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index adfca57420d1..3a898712d6cd 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -1151,6 +1151,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
 		vcpu->arch.gprs[i] = regs->gpr[i];
 	vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
@@ -1158,6 +1160,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	vcpu->arch.lo = regs->lo;
 	vcpu->arch.pc = regs->pc;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index d85bfd733ccd..24bc7aabfc44 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -528,6 +528,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	kvmppc_set_pc(vcpu, regs->pc);
 	kvmppc_set_cr(vcpu, regs->cr);
 	kvmppc_set_ctr(vcpu, regs->ctr);
@@ -548,6 +550,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index e0e4f04c5535..bcbbeddc3430 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1462,6 +1462,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	vcpu->arch.pc = regs->pc;
 	kvmppc_set_cr(vcpu, regs->cr);
 	vcpu->arch.ctr = regs->ctr;
@@ -1483,6 +1485,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 37b7caae2484..e3476430578a 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2712,7 +2712,9 @@ static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
 	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 597e1f8fc8da..75eacce78f59 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7350,6 +7350,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
+
 	vcpu->arch.emulate_regs_need_sync_from_vcpu = true;
 	vcpu->arch.emulate_regs_need_sync_to_vcpu = false;
 
@@ -7379,6 +7381,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 843d481f58cb..963e249d7b79 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2572,9 +2572,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			r = PTR_ERR(kvm_regs);
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
-		vcpu_put(vcpu);
 		kfree(kvm_regs);
 		break;
 	}
-- 
2.14.2

[PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c      | 3 +++
 arch/powerpc/kvm/book3s.c | 3 +++
 arch/powerpc/kvm/booke.c  | 3 +++
 arch/s390/kvm/kvm-s390.c  | 2 ++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 6 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index adfca57420d1..3a898712d6cd 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -1151,6 +1151,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
 		vcpu->arch.gprs[i] = regs->gpr[i];
 	vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
@@ -1158,6 +1160,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	vcpu->arch.lo = regs->lo;
 	vcpu->arch.pc = regs->pc;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index d85bfd733ccd..24bc7aabfc44 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -528,6 +528,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	kvmppc_set_pc(vcpu, regs->pc);
 	kvmppc_set_cr(vcpu, regs->cr);
 	kvmppc_set_ctr(vcpu, regs->ctr);
@@ -548,6 +550,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index e0e4f04c5535..bcbbeddc3430 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1462,6 +1462,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	vcpu->arch.pc = regs->pc;
 	kvmppc_set_cr(vcpu, regs->cr);
 	vcpu->arch.ctr = regs->ctr;
@@ -1483,6 +1485,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 37b7caae2484..e3476430578a 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2712,7 +2712,9 @@ static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
 	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 597e1f8fc8da..75eacce78f59 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7350,6 +7350,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
+
 	vcpu->arch.emulate_regs_need_sync_from_vcpu = true;
 	vcpu->arch.emulate_regs_need_sync_to_vcpu = false;
 
@@ -7379,6 +7381,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 843d481f58cb..963e249d7b79 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2572,9 +2572,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			r = PTR_ERR(kvm_regs);
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
-		vcpu_put(vcpu);
 		kfree(kvm_regs);
 		break;
 	}
-- 
2.14.2

[PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c      | 3 +++
 arch/powerpc/kvm/book3s.c | 3 +++
 arch/powerpc/kvm/booke.c  | 3 +++
 arch/s390/kvm/kvm-s390.c  | 2 ++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 6 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index adfca57420d1..3a898712d6cd 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -1151,6 +1151,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
 		vcpu->arch.gprs[i] = regs->gpr[i];
 	vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
@@ -1158,6 +1160,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	vcpu->arch.lo = regs->lo;
 	vcpu->arch.pc = regs->pc;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index d85bfd733ccd..24bc7aabfc44 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -528,6 +528,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	kvmppc_set_pc(vcpu, regs->pc);
 	kvmppc_set_cr(vcpu, regs->cr);
 	kvmppc_set_ctr(vcpu, regs->ctr);
@@ -548,6 +550,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index e0e4f04c5535..bcbbeddc3430 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1462,6 +1462,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	vcpu->arch.pc = regs->pc;
 	kvmppc_set_cr(vcpu, regs->cr);
 	vcpu->arch.ctr = regs->ctr;
@@ -1483,6 +1485,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 37b7caae2484..e3476430578a 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2712,7 +2712,9 @@ static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
 	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 597e1f8fc8da..75eacce78f59 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7350,6 +7350,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
+
 	vcpu->arch.emulate_regs_need_sync_from_vcpu = true;
 	vcpu->arch.emulate_regs_need_sync_to_vcpu = false;
 
@@ -7379,6 +7381,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 843d481f58cb..963e249d7b79 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2572,9 +2572,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			r = PTR_ERR(kvm_regs);
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
-		vcpu_put(vcpu);
 		kfree(kvm_regs);
 		break;
 	}
-- 
2.14.2

[PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_regs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c      | 3 +++
 arch/powerpc/kvm/book3s.c | 3 +++
 arch/powerpc/kvm/booke.c  | 3 +++
 arch/s390/kvm/kvm-s390.c  | 2 ++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 6 files changed, 14 insertions(+), 2 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index adfca57420d1..3a898712d6cd 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -1151,6 +1151,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
 		vcpu->arch.gprs[i] = regs->gpr[i];
 	vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
@@ -1158,6 +1160,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	vcpu->arch.lo = regs->lo;
 	vcpu->arch.pc = regs->pc;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index d85bfd733ccd..24bc7aabfc44 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -528,6 +528,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	kvmppc_set_pc(vcpu, regs->pc);
 	kvmppc_set_cr(vcpu, regs->cr);
 	kvmppc_set_ctr(vcpu, regs->ctr);
@@ -548,6 +550,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index e0e4f04c5535..bcbbeddc3430 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1462,6 +1462,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
 	int i;
 
+	vcpu_load(vcpu);
+
 	vcpu->arch.pc = regs->pc;
 	kvmppc_set_cr(vcpu, regs->cr);
 	vcpu->arch.ctr = regs->ctr;
@@ -1483,6 +1485,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
 		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 37b7caae2484..e3476430578a 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2712,7 +2712,9 @@ static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
 	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 597e1f8fc8da..75eacce78f59 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7350,6 +7350,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 {
+	vcpu_load(vcpu);
+
 	vcpu->arch.emulate_regs_need_sync_from_vcpu = true;
 	vcpu->arch.emulate_regs_need_sync_to_vcpu = false;
 
@@ -7379,6 +7381,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 843d481f58cb..963e249d7b79 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2572,9 +2572,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			r = PTR_ERR(kvm_regs);
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
-		vcpu_put(vcpu);
 		kfree(kvm_regs);
 		break;
 	}
-- 
2.14.2

Re: [PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index adfca57420d1..3a898712d6cd 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -1151,6 +1151,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
>  		vcpu->arch.gprs[i] = regs->gpr[i];
>  	vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
> @@ -1158,6 +1160,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	vcpu->arch.lo = regs->lo;
>  	vcpu->arch.pc = regs->pc;
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index d85bfd733ccd..24bc7aabfc44 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -528,6 +528,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	kvmppc_set_pc(vcpu, regs->pc);
>  	kvmppc_set_cr(vcpu, regs->cr);
>  	kvmppc_set_ctr(vcpu, regs->ctr);
> @@ -548,6 +550,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index e0e4f04c5535..bcbbeddc3430 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1462,6 +1462,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	vcpu->arch.pc = regs->pc;
>  	kvmppc_set_cr(vcpu, regs->cr);
>  	vcpu->arch.ctr = regs->ctr;
> @@ -1483,6 +1485,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 37b7caae2484..e3476430578a 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2712,7 +2712,9 @@ static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
>  
>  int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
>  	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 597e1f8fc8da..75eacce78f59 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7350,6 +7350,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
> +
>  	vcpu->arch.emulate_regs_need_sync_from_vcpu = true;
>  	vcpu->arch.emulate_regs_need_sync_to_vcpu = false;
>  
> @@ -7379,6 +7381,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  	kvm_make_request(KVM_REQ_EVENT, vcpu);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 843d481f58cb..963e249d7b79 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2572,9 +2572,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  			r = PTR_ERR(kvm_regs);
>  			goto out;
>  		}
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
> -		vcpu_put(vcpu);
>  		kfree(kvm_regs);
>  		break;
>  	}
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index adfca57420d1..3a898712d6cd 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -1151,6 +1151,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
>  		vcpu->arch.gprs[i] = regs->gpr[i];
>  	vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
> @@ -1158,6 +1160,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	vcpu->arch.lo = regs->lo;
>  	vcpu->arch.pc = regs->pc;
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index d85bfd733ccd..24bc7aabfc44 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -528,6 +528,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	kvmppc_set_pc(vcpu, regs->pc);
>  	kvmppc_set_cr(vcpu, regs->cr);
>  	kvmppc_set_ctr(vcpu, regs->ctr);
> @@ -548,6 +550,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index e0e4f04c5535..bcbbeddc3430 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1462,6 +1462,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	vcpu->arch.pc = regs->pc;
>  	kvmppc_set_cr(vcpu, regs->cr);
>  	vcpu->arch.ctr = regs->ctr;
> @@ -1483,6 +1485,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 37b7caae2484..e3476430578a 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2712,7 +2712,9 @@ static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
>  
>  int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
>  	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 597e1f8fc8da..75eacce78f59 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7350,6 +7350,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
> +
>  	vcpu->arch.emulate_regs_need_sync_from_vcpu = true;
>  	vcpu->arch.emulate_regs_need_sync_to_vcpu = false;
>  
> @@ -7379,6 +7381,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  	kvm_make_request(KVM_REQ_EVENT, vcpu);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 843d481f58cb..963e249d7b79 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2572,9 +2572,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  			r = PTR_ERR(kvm_regs);
>  			goto out;
>  		}
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
> -		vcpu_put(vcpu);
>  		kfree(kvm_regs);
>  		break;
>  	}
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

[PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index adfca57420d1..3a898712d6cd 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -1151,6 +1151,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
>  		vcpu->arch.gprs[i] = regs->gpr[i];
>  	vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
> @@ -1158,6 +1160,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	vcpu->arch.lo = regs->lo;
>  	vcpu->arch.pc = regs->pc;
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index d85bfd733ccd..24bc7aabfc44 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -528,6 +528,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	kvmppc_set_pc(vcpu, regs->pc);
>  	kvmppc_set_cr(vcpu, regs->cr);
>  	kvmppc_set_ctr(vcpu, regs->ctr);
> @@ -548,6 +550,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index e0e4f04c5535..bcbbeddc3430 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1462,6 +1462,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	vcpu->arch.pc = regs->pc;
>  	kvmppc_set_cr(vcpu, regs->cr);
>  	vcpu->arch.ctr = regs->ctr;
> @@ -1483,6 +1485,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 37b7caae2484..e3476430578a 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2712,7 +2712,9 @@ static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
>  
>  int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
>  	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 597e1f8fc8da..75eacce78f59 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7350,6 +7350,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
> +
>  	vcpu->arch.emulate_regs_need_sync_from_vcpu = true;
>  	vcpu->arch.emulate_regs_need_sync_to_vcpu = false;
>  
> @@ -7379,6 +7381,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  	kvm_make_request(KVM_REQ_EVENT, vcpu);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 843d481f58cb..963e249d7b79 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2572,9 +2572,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  			r = PTR_ERR(kvm_regs);
>  			goto out;
>  		}
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
> -		vcpu_put(vcpu);
>  		kfree(kvm_regs);
>  		break;
>  	}
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index adfca57420d1..3a898712d6cd 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -1151,6 +1151,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	for (i = 1; i < ARRAY_SIZE(vcpu->arch.gprs); i++)
>  		vcpu->arch.gprs[i] = regs->gpr[i];
>  	vcpu->arch.gprs[0] = 0; /* zero is special, and cannot be set. */
> @@ -1158,6 +1160,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	vcpu->arch.lo = regs->lo;
>  	vcpu->arch.pc = regs->pc;
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index d85bfd733ccd..24bc7aabfc44 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -528,6 +528,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	kvmppc_set_pc(vcpu, regs->pc);
>  	kvmppc_set_cr(vcpu, regs->cr);
>  	kvmppc_set_ctr(vcpu, regs->ctr);
> @@ -548,6 +550,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index e0e4f04c5535..bcbbeddc3430 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1462,6 +1462,8 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
>  	int i;
>  
> +	vcpu_load(vcpu);
> +
>  	vcpu->arch.pc = regs->pc;
>  	kvmppc_set_cr(vcpu, regs->cr);
>  	vcpu->arch.ctr = regs->ctr;
> @@ -1483,6 +1485,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  	for (i = 0; i < ARRAY_SIZE(regs->gpr); i++)
>  		kvmppc_set_gpr(vcpu, i, regs->gpr[i]);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 37b7caae2484..e3476430578a 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2712,7 +2712,9 @@ static int kvm_arch_vcpu_ioctl_initial_reset(struct kvm_vcpu *vcpu)
>  
>  int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
>  	memcpy(&vcpu->run->s.regs.gprs, &regs->gprs, sizeof(regs->gprs));
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 597e1f8fc8da..75eacce78f59 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7350,6 +7350,8 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  {
> +	vcpu_load(vcpu);
> +
>  	vcpu->arch.emulate_regs_need_sync_from_vcpu = true;
>  	vcpu->arch.emulate_regs_need_sync_to_vcpu = false;
>  
> @@ -7379,6 +7381,7 @@ int kvm_arch_vcpu_ioctl_set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  
>  	kvm_make_request(KVM_REQ_EVENT, vcpu);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 843d481f58cb..963e249d7b79 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2572,9 +2572,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  			r = PTR_ERR(kvm_regs);
>  			goto out;
>  		}
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_set_regs(vcpu, kvm_regs);
> -		vcpu_put(vcpu);
>  		kfree(kvm_regs);
>  		break;
>  	}
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:27 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:27 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 05/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_regs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:27 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_regs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c      | 3 +++
>  arch/powerpc/kvm/book3s.c | 3 +++
>  arch/powerpc/kvm/booke.c  | 3 +++
>  arch/s390/kvm/kvm-s390.c  | 2 ++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  6 files changed, 14 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/book3s.c | 8 +++++++-
 arch/powerpc/kvm/booke.c  | 9 ++++++++-
 arch/s390/kvm/kvm-s390.c  | 4 ++++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 5 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 24bc7aabfc44..6cc2377549f7 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -484,7 +484,13 @@ void kvmppc_subarch_vcpu_uninit(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
-	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	int ret;
+
+	vcpu_load(vcpu);
+	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	vcpu_put(vcpu);
+
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index bcbbeddc3430..f647e121070e 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1613,11 +1613,18 @@ int kvmppc_set_sregs_ivor(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
                                   struct kvm_sregs *sregs)
 {
+	int ret;
+
+	vcpu_load(vcpu);
+
 	sregs->pvr = vcpu->arch.pvr;
 
 	get_sregs_base(vcpu, sregs);
 	get_sregs_arch206(vcpu, sregs);
-	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index e3476430578a..18011fc4ac49 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2737,8 +2737,12 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
+	vcpu_load(vcpu);
+
 	memcpy(&sregs->acrs, &vcpu->run->s.regs.acrs, sizeof(sregs->acrs));
 	memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs));
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 75eacce78f59..20a5f6776eea 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7400,6 +7400,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 {
 	struct desc_ptr dt;
 
+	vcpu_load(vcpu);
+
 	kvm_get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
 	kvm_get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
 	kvm_get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
@@ -7431,6 +7433,7 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 		set_bit(vcpu->arch.interrupt.nr,
 			(unsigned long *)sregs->interrupt_bitmap);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 963e249d7b79..779c03e39fa4 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2581,9 +2581,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!kvm_sregs)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/book3s.c | 8 +++++++-
 arch/powerpc/kvm/booke.c  | 9 ++++++++-
 arch/s390/kvm/kvm-s390.c  | 4 ++++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 5 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 24bc7aabfc44..6cc2377549f7 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -484,7 +484,13 @@ void kvmppc_subarch_vcpu_uninit(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
-	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	int ret;
+
+	vcpu_load(vcpu);
+	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	vcpu_put(vcpu);
+
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index bcbbeddc3430..f647e121070e 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1613,11 +1613,18 @@ int kvmppc_set_sregs_ivor(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
                                   struct kvm_sregs *sregs)
 {
+	int ret;
+
+	vcpu_load(vcpu);
+
 	sregs->pvr = vcpu->arch.pvr;
 
 	get_sregs_base(vcpu, sregs);
 	get_sregs_arch206(vcpu, sregs);
-	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index e3476430578a..18011fc4ac49 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2737,8 +2737,12 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
+	vcpu_load(vcpu);
+
 	memcpy(&sregs->acrs, &vcpu->run->s.regs.acrs, sizeof(sregs->acrs));
 	memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs));
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 75eacce78f59..20a5f6776eea 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7400,6 +7400,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 {
 	struct desc_ptr dt;
 
+	vcpu_load(vcpu);
+
 	kvm_get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
 	kvm_get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
 	kvm_get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
@@ -7431,6 +7433,7 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 		set_bit(vcpu->arch.interrupt.nr,
 			(unsigned long *)sregs->interrupt_bitmap);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 963e249d7b79..779c03e39fa4 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2581,9 +2581,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!kvm_sregs)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/book3s.c | 8 +++++++-
 arch/powerpc/kvm/booke.c  | 9 ++++++++-
 arch/s390/kvm/kvm-s390.c  | 4 ++++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 5 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 24bc7aabfc44..6cc2377549f7 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -484,7 +484,13 @@ void kvmppc_subarch_vcpu_uninit(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
-	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	int ret;
+
+	vcpu_load(vcpu);
+	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	vcpu_put(vcpu);
+
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index bcbbeddc3430..f647e121070e 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1613,11 +1613,18 @@ int kvmppc_set_sregs_ivor(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
                                   struct kvm_sregs *sregs)
 {
+	int ret;
+
+	vcpu_load(vcpu);
+
 	sregs->pvr = vcpu->arch.pvr;
 
 	get_sregs_base(vcpu, sregs);
 	get_sregs_arch206(vcpu, sregs);
-	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index e3476430578a..18011fc4ac49 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2737,8 +2737,12 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
+	vcpu_load(vcpu);
+
 	memcpy(&sregs->acrs, &vcpu->run->s.regs.acrs, sizeof(sregs->acrs));
 	memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs));
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 75eacce78f59..20a5f6776eea 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7400,6 +7400,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 {
 	struct desc_ptr dt;
 
+	vcpu_load(vcpu);
+
 	kvm_get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
 	kvm_get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
 	kvm_get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
@@ -7431,6 +7433,7 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 		set_bit(vcpu->arch.interrupt.nr,
 			(unsigned long *)sregs->interrupt_bitmap);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 963e249d7b79..779c03e39fa4 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2581,9 +2581,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!kvm_sregs)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/book3s.c | 8 +++++++-
 arch/powerpc/kvm/booke.c  | 9 ++++++++-
 arch/s390/kvm/kvm-s390.c  | 4 ++++
 arch/x86/kvm/x86.c        | 3 +++
 virt/kvm/kvm_main.c       | 2 --
 5 files changed, 22 insertions(+), 4 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 24bc7aabfc44..6cc2377549f7 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -484,7 +484,13 @@ void kvmppc_subarch_vcpu_uninit(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
-	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	int ret;
+
+	vcpu_load(vcpu);
+	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	vcpu_put(vcpu);
+
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index bcbbeddc3430..f647e121070e 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1613,11 +1613,18 @@ int kvmppc_set_sregs_ivor(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
                                   struct kvm_sregs *sregs)
 {
+	int ret;
+
+	vcpu_load(vcpu);
+
 	sregs->pvr = vcpu->arch.pvr;
 
 	get_sregs_base(vcpu, sregs);
 	get_sregs_arch206(vcpu, sregs);
-	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
+
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index e3476430578a..18011fc4ac49 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2737,8 +2737,12 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
+	vcpu_load(vcpu);
+
 	memcpy(&sregs->acrs, &vcpu->run->s.regs.acrs, sizeof(sregs->acrs));
 	memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs));
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 75eacce78f59..20a5f6776eea 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7400,6 +7400,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 {
 	struct desc_ptr dt;
 
+	vcpu_load(vcpu);
+
 	kvm_get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
 	kvm_get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
 	kvm_get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
@@ -7431,6 +7433,7 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 		set_bit(vcpu->arch.interrupt.nr,
 			(unsigned long *)sregs->interrupt_bitmap);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 963e249d7b79..779c03e39fa4 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2581,9 +2581,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!kvm_sregs)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

Re: [PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c | 8 +++++++-
>  arch/powerpc/kvm/booke.c  | 9 ++++++++-
>  arch/s390/kvm/kvm-s390.c  | 4 ++++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  5 files changed, 22 insertions(+), 4 deletions(-)
> 
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index 24bc7aabfc44..6cc2377549f7 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -484,7 +484,13 @@ void kvmppc_subarch_vcpu_uninit(struct kvm_vcpu *vcpu)
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> -	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	int ret;
> +
> +	vcpu_load(vcpu);
> +	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	vcpu_put(vcpu);
> +
> +	return ret;
>  }
>  
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index bcbbeddc3430..f647e121070e 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1613,11 +1613,18 @@ int kvmppc_set_sregs_ivor(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>                                    struct kvm_sregs *sregs)
>  {
> +	int ret;
> +
> +	vcpu_load(vcpu);
> +
>  	sregs->pvr = vcpu->arch.pvr;
>  
>  	get_sregs_base(vcpu, sregs);
>  	get_sregs_arch206(vcpu, sregs);
> -	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index e3476430578a..18011fc4ac49 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2737,8 +2737,12 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> +	vcpu_load(vcpu);
> +
>  	memcpy(&sregs->acrs, &vcpu->run->s.regs.acrs, sizeof(sregs->acrs));
>  	memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs));
> +
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 75eacce78f59..20a5f6776eea 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7400,6 +7400,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  {
>  	struct desc_ptr dt;
>  
> +	vcpu_load(vcpu);
> +
>  	kvm_get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
>  	kvm_get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
>  	kvm_get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
> @@ -7431,6 +7433,7 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  		set_bit(vcpu->arch.interrupt.nr,
>  			(unsigned long *)sregs->interrupt_bitmap);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 963e249d7b79..779c03e39fa4 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2581,9 +2581,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		r = -ENOMEM;
>  		if (!kvm_sregs)
>  			goto out;
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
> -		vcpu_put(vcpu);
>  		if (r)
>  			goto out;
>  		r = -EFAULT;
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c | 8 +++++++-
>  arch/powerpc/kvm/booke.c  | 9 ++++++++-
>  arch/s390/kvm/kvm-s390.c  | 4 ++++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  5 files changed, 22 insertions(+), 4 deletions(-)
> 
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index 24bc7aabfc44..6cc2377549f7 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -484,7 +484,13 @@ void kvmppc_subarch_vcpu_uninit(struct kvm_vcpu *vcpu)
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> -	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	int ret;
> +
> +	vcpu_load(vcpu);
> +	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	vcpu_put(vcpu);
> +
> +	return ret;
>  }
>  
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index bcbbeddc3430..f647e121070e 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1613,11 +1613,18 @@ int kvmppc_set_sregs_ivor(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>                                    struct kvm_sregs *sregs)
>  {
> +	int ret;
> +
> +	vcpu_load(vcpu);
> +
>  	sregs->pvr = vcpu->arch.pvr;
>  
>  	get_sregs_base(vcpu, sregs);
>  	get_sregs_arch206(vcpu, sregs);
> -	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index e3476430578a..18011fc4ac49 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2737,8 +2737,12 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> +	vcpu_load(vcpu);
> +
>  	memcpy(&sregs->acrs, &vcpu->run->s.regs.acrs, sizeof(sregs->acrs));
>  	memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs));
> +
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 75eacce78f59..20a5f6776eea 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7400,6 +7400,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  {
>  	struct desc_ptr dt;
>  
> +	vcpu_load(vcpu);
> +
>  	kvm_get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
>  	kvm_get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
>  	kvm_get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
> @@ -7431,6 +7433,7 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  		set_bit(vcpu->arch.interrupt.nr,
>  			(unsigned long *)sregs->interrupt_bitmap);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 963e249d7b79..779c03e39fa4 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2581,9 +2581,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		r = -ENOMEM;
>  		if (!kvm_sregs)
>  			goto out;
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
> -		vcpu_put(vcpu);
>  		if (r)
>  			goto out;
>  		r = -EFAULT;
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

[PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[David Hildenbrand]

On 04.12.2017 21:35, Christoffer Dall wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c | 8 +++++++-
>  arch/powerpc/kvm/booke.c  | 9 ++++++++-
>  arch/s390/kvm/kvm-s390.c  | 4 ++++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  5 files changed, 22 insertions(+), 4 deletions(-)
> 
> diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
> index 24bc7aabfc44..6cc2377549f7 100644
> --- a/arch/powerpc/kvm/book3s.c
> +++ b/arch/powerpc/kvm/book3s.c
> @@ -484,7 +484,13 @@ void kvmppc_subarch_vcpu_uninit(struct kvm_vcpu *vcpu)
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> -	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	int ret;
> +
> +	vcpu_load(vcpu);
> +	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	vcpu_put(vcpu);
> +
> +	return ret;
>  }
>  
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index bcbbeddc3430..f647e121070e 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1613,11 +1613,18 @@ int kvmppc_set_sregs_ivor(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs)
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>                                    struct kvm_sregs *sregs)
>  {
> +	int ret;
> +
> +	vcpu_load(vcpu);
> +
>  	sregs->pvr = vcpu->arch.pvr;
>  
>  	get_sregs_base(vcpu, sregs);
>  	get_sregs_arch206(vcpu, sregs);
> -	return vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +	ret = vcpu->kvm->arch.kvm_ops->get_sregs(vcpu, sregs);
> +
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index e3476430578a..18011fc4ac49 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2737,8 +2737,12 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> +	vcpu_load(vcpu);
> +
>  	memcpy(&sregs->acrs, &vcpu->run->s.regs.acrs, sizeof(sregs->acrs));
>  	memcpy(&sregs->crs, &vcpu->arch.sie_block->gcr, sizeof(sregs->crs));
> +
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 75eacce78f59..20a5f6776eea 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7400,6 +7400,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  {
>  	struct desc_ptr dt;
>  
> +	vcpu_load(vcpu);
> +
>  	kvm_get_segment(vcpu, &sregs->cs, VCPU_SREG_CS);
>  	kvm_get_segment(vcpu, &sregs->ds, VCPU_SREG_DS);
>  	kvm_get_segment(vcpu, &sregs->es, VCPU_SREG_ES);
> @@ -7431,6 +7433,7 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
>  		set_bit(vcpu->arch.interrupt.nr,
>  			(unsigned long *)sregs->interrupt_bitmap);
>  
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 963e249d7b79..779c03e39fa4 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2581,9 +2581,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		r = -ENOMEM;
>  		if (!kvm_sregs)
>  			goto out;
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl_get_sregs(vcpu, kvm_sregs);
> -		vcpu_put(vcpu);
>  		if (r)
>  			goto out;
>  		r = -EFAULT;
> 

Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:28 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c | 8 +++++++-
>  arch/powerpc/kvm/booke.c  | 9 ++++++++-
>  arch/s390/kvm/kvm-s390.c  | 4 ++++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  5 files changed, 22 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:28 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c | 8 +++++++-
>  arch/powerpc/kvm/booke.c  | 9 ++++++++-
>  arch/s390/kvm/kvm-s390.c  | 4 ++++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  5 files changed, 22 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 06/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_sregs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:28 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c | 8 +++++++-
>  arch/powerpc/kvm/booke.c  | 9 ++++++++-
>  arch/s390/kvm/kvm-s390.c  | 4 ++++
>  arch/x86/kvm/x86.c        | 3 +++
>  virt/kvm/kvm_main.c       | 2 --
>  5 files changed, 22 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/book3s.c |  8 +++++++-
 arch/powerpc/kvm/booke.c  | 15 +++++++++++----
 arch/s390/kvm/kvm-s390.c  |  4 ++++
 arch/x86/kvm/x86.c        | 13 ++++++++++---
 virt/kvm/kvm_main.c       |  2 --
 5 files changed, 32 insertions(+), 10 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 6cc2377549f7..047651622cb8 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -496,7 +496,13 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
-	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+	int ret;
+
+	vcpu_load(vcpu);
+	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+	vcpu_put(vcpu);
+
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index f647e121070e..cdf0be02c95a 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 {
 	int ret;
 
+	vcpu_load(vcpu);
+
+	ret = -EINVAL;
 	if (vcpu->arch.pvr != sregs->pvr)
-		return -EINVAL;
+		goto out;
 
 	ret = set_sregs_base(vcpu, sregs);
 	if (ret < 0)
-		return ret;
+		goto out;
 
 	ret = set_sregs_arch206(vcpu, sregs);
 	if (ret < 0)
-		return ret;
+		goto out;
+
+	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
 
-	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 18011fc4ac49..d95b4f15e52b 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
+	vcpu_load(vcpu);
+
 	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
 	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 20a5f6776eea..a31a80aee0b9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	int mmu_reset_needed = 0;
 	int pending_vec, max_bits, idx;
 	struct desc_ptr dt;
+	int ret;
+
+	vcpu_load(vcpu);
 
+	ret = -EINVAL;
 	if (!guest_cpuid_has(vcpu, X86_FEATURE_XSAVE) &&
 			(sregs->cr4 & X86_CR4_OSXSAVE))
-		return -EINVAL;
+		goto out;
 
 	apic_base_msr.data = sregs->apic_base;
 	apic_base_msr.host_initiated = true;
 	if (kvm_set_apic_base(vcpu, &apic_base_msr))
-		return -EINVAL;
+		goto out;
 
 	dt.size = sregs->idt.limit;
 	dt.address = sregs->idt.base;
@@ -7574,7 +7578,10 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
-	return 0;
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 779c03e39fa4..19cf2d11f80f 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2597,9 +2597,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			kvm_sregs = NULL;
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_sregs(vcpu, kvm_sregs);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_GET_MP_STATE: {
-- 
2.14.2

[PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/book3s.c |  8 +++++++-
 arch/powerpc/kvm/booke.c  | 15 +++++++++++----
 arch/s390/kvm/kvm-s390.c  |  4 ++++
 arch/x86/kvm/x86.c        | 13 ++++++++++---
 virt/kvm/kvm_main.c       |  2 --
 5 files changed, 32 insertions(+), 10 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 6cc2377549f7..047651622cb8 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -496,7 +496,13 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
-	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+	int ret;
+
+	vcpu_load(vcpu);
+	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+	vcpu_put(vcpu);
+
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index f647e121070e..cdf0be02c95a 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 {
 	int ret;
 
+	vcpu_load(vcpu);
+
+	ret = -EINVAL;
 	if (vcpu->arch.pvr != sregs->pvr)
-		return -EINVAL;
+		goto out;
 
 	ret = set_sregs_base(vcpu, sregs);
 	if (ret < 0)
-		return ret;
+		goto out;
 
 	ret = set_sregs_arch206(vcpu, sregs);
 	if (ret < 0)
-		return ret;
+		goto out;
+
+	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
 
-	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 18011fc4ac49..d95b4f15e52b 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
+	vcpu_load(vcpu);
+
 	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
 	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 20a5f6776eea..a31a80aee0b9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	int mmu_reset_needed = 0;
 	int pending_vec, max_bits, idx;
 	struct desc_ptr dt;
+	int ret;
+
+	vcpu_load(vcpu);
 
+	ret = -EINVAL;
 	if (!guest_cpuid_has(vcpu, X86_FEATURE_XSAVE) &&
 			(sregs->cr4 & X86_CR4_OSXSAVE))
-		return -EINVAL;
+		goto out;
 
 	apic_base_msr.data = sregs->apic_base;
 	apic_base_msr.host_initiated = true;
 	if (kvm_set_apic_base(vcpu, &apic_base_msr))
-		return -EINVAL;
+		goto out;
 
 	dt.size = sregs->idt.limit;
 	dt.address = sregs->idt.base;
@@ -7574,7 +7578,10 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
-	return 0;
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 779c03e39fa4..19cf2d11f80f 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2597,9 +2597,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			kvm_sregs = NULL;
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_sregs(vcpu, kvm_sregs);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_GET_MP_STATE: {
-- 
2.14.2

[PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/book3s.c |  8 +++++++-
 arch/powerpc/kvm/booke.c  | 15 +++++++++++----
 arch/s390/kvm/kvm-s390.c  |  4 ++++
 arch/x86/kvm/x86.c        | 13 ++++++++++---
 virt/kvm/kvm_main.c       |  2 --
 5 files changed, 32 insertions(+), 10 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 6cc2377549f7..047651622cb8 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -496,7 +496,13 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
-	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+	int ret;
+
+	vcpu_load(vcpu);
+	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+	vcpu_put(vcpu);
+
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index f647e121070e..cdf0be02c95a 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 {
 	int ret;
 
+	vcpu_load(vcpu);
+
+	ret = -EINVAL;
 	if (vcpu->arch.pvr != sregs->pvr)
-		return -EINVAL;
+		goto out;
 
 	ret = set_sregs_base(vcpu, sregs);
 	if (ret < 0)
-		return ret;
+		goto out;
 
 	ret = set_sregs_arch206(vcpu, sregs);
 	if (ret < 0)
-		return ret;
+		goto out;
+
+	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
 
-	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 18011fc4ac49..d95b4f15e52b 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
+	vcpu_load(vcpu);
+
 	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
 	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 20a5f6776eea..a31a80aee0b9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	int mmu_reset_needed = 0;
 	int pending_vec, max_bits, idx;
 	struct desc_ptr dt;
+	int ret;
+
+	vcpu_load(vcpu);
 
+	ret = -EINVAL;
 	if (!guest_cpuid_has(vcpu, X86_FEATURE_XSAVE) &&
 			(sregs->cr4 & X86_CR4_OSXSAVE))
-		return -EINVAL;
+		goto out;
 
 	apic_base_msr.data = sregs->apic_base;
 	apic_base_msr.host_initiated = true;
 	if (kvm_set_apic_base(vcpu, &apic_base_msr))
-		return -EINVAL;
+		goto out;
 
 	dt.size = sregs->idt.limit;
 	dt.address = sregs->idt.base;
@@ -7574,7 +7578,10 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
-	return 0;
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 779c03e39fa4..19cf2d11f80f 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2597,9 +2597,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			kvm_sregs = NULL;
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_sregs(vcpu, kvm_sregs);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_GET_MP_STATE: {
-- 
2.14.2

[PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_sregs().

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/book3s.c |  8 +++++++-
 arch/powerpc/kvm/booke.c  | 15 +++++++++++----
 arch/s390/kvm/kvm-s390.c  |  4 ++++
 arch/x86/kvm/x86.c        | 13 ++++++++++---
 virt/kvm/kvm_main.c       |  2 --
 5 files changed, 32 insertions(+), 10 deletions(-)

diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 6cc2377549f7..047651622cb8 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -496,7 +496,13 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
-	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+	int ret;
+
+	vcpu_load(vcpu);
+	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+	vcpu_put(vcpu);
+
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index f647e121070e..cdf0be02c95a 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 {
 	int ret;
 
+	vcpu_load(vcpu);
+
+	ret = -EINVAL;
 	if (vcpu->arch.pvr != sregs->pvr)
-		return -EINVAL;
+		goto out;
 
 	ret = set_sregs_base(vcpu, sregs);
 	if (ret < 0)
-		return ret;
+		goto out;
 
 	ret = set_sregs_arch206(vcpu, sregs);
 	if (ret < 0)
-		return ret;
+		goto out;
+
+	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
 
-	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 18011fc4ac49..d95b4f15e52b 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
 int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 				  struct kvm_sregs *sregs)
 {
+	vcpu_load(vcpu);
+
 	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
 	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 20a5f6776eea..a31a80aee0b9 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 	int mmu_reset_needed = 0;
 	int pending_vec, max_bits, idx;
 	struct desc_ptr dt;
+	int ret;
+
+	vcpu_load(vcpu);
 
+	ret = -EINVAL;
 	if (!guest_cpuid_has(vcpu, X86_FEATURE_XSAVE) &&
 			(sregs->cr4 & X86_CR4_OSXSAVE))
-		return -EINVAL;
+		goto out;
 
 	apic_base_msr.data = sregs->apic_base;
 	apic_base_msr.host_initiated = true;
 	if (kvm_set_apic_base(vcpu, &apic_base_msr))
-		return -EINVAL;
+		goto out;
 
 	dt.size = sregs->idt.limit;
 	dt.address = sregs->idt.base;
@@ -7574,7 +7578,10 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
 
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
 
-	return 0;
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 779c03e39fa4..19cf2d11f80f 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2597,9 +2597,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			kvm_sregs = NULL;
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_sregs(vcpu, kvm_sregs);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_GET_MP_STATE: {
-- 
2.14.2

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[David Hildenbrand]

>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index f647e121070e..cdf0be02c95a 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  {
>  	int ret;
>  
> +	vcpu_load(vcpu);
> +
> +	ret = -EINVAL;

you can initialize this directly.

>  	if (vcpu->arch.pvr != sregs->pvr)
> -		return -EINVAL;
> +		goto out;
>  
>  	ret = set_sregs_base(vcpu, sregs);
>  	if (ret < 0)
> -		return ret;
> +		goto out;
>  
>  	ret = set_sregs_arch206(vcpu, sregs);
>  	if (ret < 0)
> -		return ret;
> +		goto out;
> +
> +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>  
> -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> +out:
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 18011fc4ac49..d95b4f15e52b 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> +	vcpu_load(vcpu);
> +
>  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
>  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
> +
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 20a5f6776eea..a31a80aee0b9 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  	int mmu_reset_needed = 0;
>  	int pending_vec, max_bits, idx;
>  	struct desc_ptr dt;
> +	int ret;
> +
> +	vcpu_load(vcpu);
>  
> +	ret = -EINVAL;

dito


Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[David Hildenbrand]

>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index f647e121070e..cdf0be02c95a 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  {
>  	int ret;
>  
> +	vcpu_load(vcpu);
> +
> +	ret = -EINVAL;

you can initialize this directly.

>  	if (vcpu->arch.pvr != sregs->pvr)
> -		return -EINVAL;
> +		goto out;
>  
>  	ret = set_sregs_base(vcpu, sregs);
>  	if (ret < 0)
> -		return ret;
> +		goto out;
>  
>  	ret = set_sregs_arch206(vcpu, sregs);
>  	if (ret < 0)
> -		return ret;
> +		goto out;
> +
> +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>  
> -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> +out:
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 18011fc4ac49..d95b4f15e52b 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> +	vcpu_load(vcpu);
> +
>  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
>  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
> +
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 20a5f6776eea..a31a80aee0b9 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  	int mmu_reset_needed = 0;
>  	int pending_vec, max_bits, idx;
>  	struct desc_ptr dt;
> +	int ret;
> +
> +	vcpu_load(vcpu);
>  
> +	ret = -EINVAL;

dito


Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

[PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[David Hildenbrand]

>  
>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index f647e121070e..cdf0be02c95a 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  {
>  	int ret;
>  
> +	vcpu_load(vcpu);
> +
> +	ret = -EINVAL;

you can initialize this directly.

>  	if (vcpu->arch.pvr != sregs->pvr)
> -		return -EINVAL;
> +		goto out;
>  
>  	ret = set_sregs_base(vcpu, sregs);
>  	if (ret < 0)
> -		return ret;
> +		goto out;
>  
>  	ret = set_sregs_arch206(vcpu, sregs);
>  	if (ret < 0)
> -		return ret;
> +		goto out;
> +
> +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>  
> -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> +out:
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 18011fc4ac49..d95b4f15e52b 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  				  struct kvm_sregs *sregs)
>  {
> +	vcpu_load(vcpu);
> +
>  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
>  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
> +
> +	vcpu_put(vcpu);
>  	return 0;
>  }
>  
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 20a5f6776eea..a31a80aee0b9 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>  	int mmu_reset_needed = 0;
>  	int pending_vec, max_bits, idx;
>  	struct desc_ptr dt;
> +	int ret;
> +
> +	vcpu_load(vcpu);
>  
> +	ret = -EINVAL;

dito


Reviewed-by: David Hildenbrand <david@redhat.com>

-- 

Thanks,

David / dhildenb

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Christoffer Dall]

On Fri, Dec 08, 2017 at 05:26:02PM +0100, David Hildenbrand wrote:
> 
> >  
> >  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> > diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> > index f647e121070e..cdf0be02c95a 100644
> > --- a/arch/powerpc/kvm/booke.c
> > +++ b/arch/powerpc/kvm/booke.c
> > @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  {
> >  	int ret;
> >  
> > +	vcpu_load(vcpu);
> > +
> > +	ret = -EINVAL;
> 
> you can initialize this directly.
> 
> >  	if (vcpu->arch.pvr != sregs->pvr)
> > -		return -EINVAL;
> > +		goto out;
> >  
> >  	ret = set_sregs_base(vcpu, sregs);
> >  	if (ret < 0)
> > -		return ret;
> > +		goto out;
> >  
> >  	ret = set_sregs_arch206(vcpu, sregs);
> >  	if (ret < 0)
> > -		return ret;
> > +		goto out;
> > +
> > +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> >  
> > -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> > +out:
> > +	vcpu_put(vcpu);
> > +	return ret;
> >  }
> >  
> >  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
> > diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> > index 18011fc4ac49..d95b4f15e52b 100644
> > --- a/arch/s390/kvm/kvm-s390.c
> > +++ b/arch/s390/kvm/kvm-s390.c
> > @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> >  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  				  struct kvm_sregs *sregs)
> >  {
> > +	vcpu_load(vcpu);
> > +
> >  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
> >  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
> > +
> > +	vcpu_put(vcpu);
> >  	return 0;
> >  }
> >  
> > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > index 20a5f6776eea..a31a80aee0b9 100644
> > --- a/arch/x86/kvm/x86.c
> > +++ b/arch/x86/kvm/x86.c
> > @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  	int mmu_reset_needed = 0;
> >  	int pending_vec, max_bits, idx;
> >  	struct desc_ptr dt;
> > +	int ret;
> > +
> > +	vcpu_load(vcpu);
> >  
> > +	ret = -EINVAL;
> 
> dito

Sure.

> 
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> 

Thanks for the review!
-Christoffer

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Christoffer Dall]

On Fri, Dec 08, 2017 at 05:26:02PM +0100, David Hildenbrand wrote:
> 
> >  
> >  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> > diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> > index f647e121070e..cdf0be02c95a 100644
> > --- a/arch/powerpc/kvm/booke.c
> > +++ b/arch/powerpc/kvm/booke.c
> > @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  {
> >  	int ret;
> >  
> > +	vcpu_load(vcpu);
> > +
> > +	ret = -EINVAL;
> 
> you can initialize this directly.
> 
> >  	if (vcpu->arch.pvr != sregs->pvr)
> > -		return -EINVAL;
> > +		goto out;
> >  
> >  	ret = set_sregs_base(vcpu, sregs);
> >  	if (ret < 0)
> > -		return ret;
> > +		goto out;
> >  
> >  	ret = set_sregs_arch206(vcpu, sregs);
> >  	if (ret < 0)
> > -		return ret;
> > +		goto out;
> > +
> > +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> >  
> > -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> > +out:
> > +	vcpu_put(vcpu);
> > +	return ret;
> >  }
> >  
> >  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
> > diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> > index 18011fc4ac49..d95b4f15e52b 100644
> > --- a/arch/s390/kvm/kvm-s390.c
> > +++ b/arch/s390/kvm/kvm-s390.c
> > @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> >  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  				  struct kvm_sregs *sregs)
> >  {
> > +	vcpu_load(vcpu);
> > +
> >  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
> >  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
> > +
> > +	vcpu_put(vcpu);
> >  	return 0;
> >  }
> >  
> > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > index 20a5f6776eea..a31a80aee0b9 100644
> > --- a/arch/x86/kvm/x86.c
> > +++ b/arch/x86/kvm/x86.c
> > @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  	int mmu_reset_needed = 0;
> >  	int pending_vec, max_bits, idx;
> >  	struct desc_ptr dt;
> > +	int ret;
> > +
> > +	vcpu_load(vcpu);
> >  
> > +	ret = -EINVAL;
> 
> dito

Sure.

> 
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> 

Thanks for the review!
-Christoffer

[PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Christoffer Dall]

On Fri, Dec 08, 2017 at 05:26:02PM +0100, David Hildenbrand wrote:
> 
> >  
> >  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> > diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> > index f647e121070e..cdf0be02c95a 100644
> > --- a/arch/powerpc/kvm/booke.c
> > +++ b/arch/powerpc/kvm/booke.c
> > @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  {
> >  	int ret;
> >  
> > +	vcpu_load(vcpu);
> > +
> > +	ret = -EINVAL;
> 
> you can initialize this directly.
> 
> >  	if (vcpu->arch.pvr != sregs->pvr)
> > -		return -EINVAL;
> > +		goto out;
> >  
> >  	ret = set_sregs_base(vcpu, sregs);
> >  	if (ret < 0)
> > -		return ret;
> > +		goto out;
> >  
> >  	ret = set_sregs_arch206(vcpu, sregs);
> >  	if (ret < 0)
> > -		return ret;
> > +		goto out;
> > +
> > +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> >  
> > -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> > +out:
> > +	vcpu_put(vcpu);
> > +	return ret;
> >  }
> >  
> >  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
> > diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> > index 18011fc4ac49..d95b4f15e52b 100644
> > --- a/arch/s390/kvm/kvm-s390.c
> > +++ b/arch/s390/kvm/kvm-s390.c
> > @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> >  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  				  struct kvm_sregs *sregs)
> >  {
> > +	vcpu_load(vcpu);
> > +
> >  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
> >  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
> > +
> > +	vcpu_put(vcpu);
> >  	return 0;
> >  }
> >  
> > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > index 20a5f6776eea..a31a80aee0b9 100644
> > --- a/arch/x86/kvm/x86.c
> > +++ b/arch/x86/kvm/x86.c
> > @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  	int mmu_reset_needed = 0;
> >  	int pending_vec, max_bits, idx;
> >  	struct desc_ptr dt;
> > +	int ret;
> > +
> > +	vcpu_load(vcpu);
> >  
> > +	ret = -EINVAL;
> 
> dito

Sure.

> 
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> 

Thanks for the review!
-Christoffer

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Christoffer Dall]

On Fri, Dec 08, 2017 at 05:26:02PM +0100, David Hildenbrand wrote:
> 
> >  
> >  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> > diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> > index f647e121070e..cdf0be02c95a 100644
> > --- a/arch/powerpc/kvm/booke.c
> > +++ b/arch/powerpc/kvm/booke.c
> > @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  {
> >  	int ret;
> >  
> > +	vcpu_load(vcpu);
> > +
> > +	ret = -EINVAL;
> 
> you can initialize this directly.
> 
> >  	if (vcpu->arch.pvr != sregs->pvr)
> > -		return -EINVAL;
> > +		goto out;
> >  
> >  	ret = set_sregs_base(vcpu, sregs);
> >  	if (ret < 0)
> > -		return ret;
> > +		goto out;
> >  
> >  	ret = set_sregs_arch206(vcpu, sregs);
> >  	if (ret < 0)
> > -		return ret;
> > +		goto out;
> > +
> > +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> >  
> > -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
> > +out:
> > +	vcpu_put(vcpu);
> > +	return ret;
> >  }
> >  
> >  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
> > diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> > index 18011fc4ac49..d95b4f15e52b 100644
> > --- a/arch/s390/kvm/kvm-s390.c
> > +++ b/arch/s390/kvm/kvm-s390.c
> > @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
> >  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  				  struct kvm_sregs *sregs)
> >  {
> > +	vcpu_load(vcpu);
> > +
> >  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
> >  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
> > +
> > +	vcpu_put(vcpu);
> >  	return 0;
> >  }
> >  
> > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > index 20a5f6776eea..a31a80aee0b9 100644
> > --- a/arch/x86/kvm/x86.c
> > +++ b/arch/x86/kvm/x86.c
> > @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
> >  	int mmu_reset_needed = 0;
> >  	int pending_vec, max_bits, idx;
> >  	struct desc_ptr dt;
> > +	int ret;
> > +
> > +	vcpu_load(vcpu);
> >  
> > +	ret = -EINVAL;
> 
> dito

Sure.

> 
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> 

Thanks for the review!
-Christoffer

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Paolo Bonzini]

On 11/12/2017 10:19, Christoffer Dall wrote:
> On Fri, Dec 08, 2017 at 05:26:02PM +0100, David Hildenbrand wrote:
>>
>>>  
>>>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>>> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
>>> index f647e121070e..cdf0be02c95a 100644
>>> --- a/arch/powerpc/kvm/booke.c
>>> +++ b/arch/powerpc/kvm/booke.c
>>> @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  {
>>>  	int ret;
>>>  
>>> +	vcpu_load(vcpu);
>>> +
>>> +	ret = -EINVAL;
>>
>> you can initialize this directly.
>>
>>>  	if (vcpu->arch.pvr != sregs->pvr)
>>> -		return -EINVAL;
>>> +		goto out;
>>>  
>>>  	ret = set_sregs_base(vcpu, sregs);
>>>  	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>>  
>>>  	ret = set_sregs_arch206(vcpu, sregs);
>>>  	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>> +
>>> +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>>>  
>>> -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>>> +out:
>>> +	vcpu_put(vcpu);
>>> +	return ret;
>>>  }
>>>  
>>>  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
>>> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
>>> index 18011fc4ac49..d95b4f15e52b 100644
>>> --- a/arch/s390/kvm/kvm-s390.c
>>> +++ b/arch/s390/kvm/kvm-s390.c
>>> @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>>>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  				  struct kvm_sregs *sregs)
>>>  {
>>> +	vcpu_load(vcpu);
>>> +
>>>  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
>>>  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
>>> +
>>> +	vcpu_put(vcpu);
>>>  	return 0;
>>>  }
>>>  
>>> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
>>> index 20a5f6776eea..a31a80aee0b9 100644
>>> --- a/arch/x86/kvm/x86.c
>>> +++ b/arch/x86/kvm/x86.c
>>> @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  	int mmu_reset_needed = 0;
>>>  	int pending_vec, max_bits, idx;
>>>  	struct desc_ptr dt;
>>> +	int ret;
>>> +
>>> +	vcpu_load(vcpu);
>>>  
>>> +	ret = -EINVAL;
>>
>> dito
> 
> Sure.

I'm doing it when applying.

Paolo

>> Reviewed-by: David Hildenbrand <david@redhat.com>

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Paolo Bonzini]

On 11/12/2017 10:19, Christoffer Dall wrote:
> On Fri, Dec 08, 2017 at 05:26:02PM +0100, David Hildenbrand wrote:
>>
>>>  
>>>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>>> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
>>> index f647e121070e..cdf0be02c95a 100644
>>> --- a/arch/powerpc/kvm/booke.c
>>> +++ b/arch/powerpc/kvm/booke.c
>>> @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  {
>>>  	int ret;
>>>  
>>> +	vcpu_load(vcpu);
>>> +
>>> +	ret = -EINVAL;
>>
>> you can initialize this directly.
>>
>>>  	if (vcpu->arch.pvr != sregs->pvr)
>>> -		return -EINVAL;
>>> +		goto out;
>>>  
>>>  	ret = set_sregs_base(vcpu, sregs);
>>>  	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>>  
>>>  	ret = set_sregs_arch206(vcpu, sregs);
>>>  	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>> +
>>> +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>>>  
>>> -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>>> +out:
>>> +	vcpu_put(vcpu);
>>> +	return ret;
>>>  }
>>>  
>>>  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
>>> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
>>> index 18011fc4ac49..d95b4f15e52b 100644
>>> --- a/arch/s390/kvm/kvm-s390.c
>>> +++ b/arch/s390/kvm/kvm-s390.c
>>> @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>>>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  				  struct kvm_sregs *sregs)
>>>  {
>>> +	vcpu_load(vcpu);
>>> +
>>>  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
>>>  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
>>> +
>>> +	vcpu_put(vcpu);
>>>  	return 0;
>>>  }
>>>  
>>> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
>>> index 20a5f6776eea..a31a80aee0b9 100644
>>> --- a/arch/x86/kvm/x86.c
>>> +++ b/arch/x86/kvm/x86.c
>>> @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  	int mmu_reset_needed = 0;
>>>  	int pending_vec, max_bits, idx;
>>>  	struct desc_ptr dt;
>>> +	int ret;
>>> +
>>> +	vcpu_load(vcpu);
>>>  
>>> +	ret = -EINVAL;
>>
>> dito
> 
> Sure.

I'm doing it when applying.

Paolo

>> Reviewed-by: David Hildenbrand <david@redhat.com>

[PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Paolo Bonzini]

On 11/12/2017 10:19, Christoffer Dall wrote:
> On Fri, Dec 08, 2017 at 05:26:02PM +0100, David Hildenbrand wrote:
>>
>>>  
>>>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>>> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
>>> index f647e121070e..cdf0be02c95a 100644
>>> --- a/arch/powerpc/kvm/booke.c
>>> +++ b/arch/powerpc/kvm/booke.c
>>> @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  {
>>>  	int ret;
>>>  
>>> +	vcpu_load(vcpu);
>>> +
>>> +	ret = -EINVAL;
>>
>> you can initialize this directly.
>>
>>>  	if (vcpu->arch.pvr != sregs->pvr)
>>> -		return -EINVAL;
>>> +		goto out;
>>>  
>>>  	ret = set_sregs_base(vcpu, sregs);
>>>  	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>>  
>>>  	ret = set_sregs_arch206(vcpu, sregs);
>>>  	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>> +
>>> +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>>>  
>>> -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>>> +out:
>>> +	vcpu_put(vcpu);
>>> +	return ret;
>>>  }
>>>  
>>>  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
>>> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
>>> index 18011fc4ac49..d95b4f15e52b 100644
>>> --- a/arch/s390/kvm/kvm-s390.c
>>> +++ b/arch/s390/kvm/kvm-s390.c
>>> @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>>>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  				  struct kvm_sregs *sregs)
>>>  {
>>> +	vcpu_load(vcpu);
>>> +
>>>  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
>>>  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
>>> +
>>> +	vcpu_put(vcpu);
>>>  	return 0;
>>>  }
>>>  
>>> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
>>> index 20a5f6776eea..a31a80aee0b9 100644
>>> --- a/arch/x86/kvm/x86.c
>>> +++ b/arch/x86/kvm/x86.c
>>> @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  	int mmu_reset_needed = 0;
>>>  	int pending_vec, max_bits, idx;
>>>  	struct desc_ptr dt;
>>> +	int ret;
>>> +
>>> +	vcpu_load(vcpu);
>>>  
>>> +	ret = -EINVAL;
>>
>> dito
> 
> Sure.

I'm doing it when applying.

Paolo

>> Reviewed-by: David Hildenbrand <david@redhat.com>

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Paolo Bonzini]

On 11/12/2017 10:19, Christoffer Dall wrote:
> On Fri, Dec 08, 2017 at 05:26:02PM +0100, David Hildenbrand wrote:
>>
>>>  
>>>  int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>>> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
>>> index f647e121070e..cdf0be02c95a 100644
>>> --- a/arch/powerpc/kvm/booke.c
>>> +++ b/arch/powerpc/kvm/booke.c
>>> @@ -1632,18 +1632,25 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  {
>>>  	int ret;
>>>  
>>> +	vcpu_load(vcpu);
>>> +
>>> +	ret = -EINVAL;
>>
>> you can initialize this directly.
>>
>>>  	if (vcpu->arch.pvr != sregs->pvr)
>>> -		return -EINVAL;
>>> +		goto out;
>>>  
>>>  	ret = set_sregs_base(vcpu, sregs);
>>>  	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>>  
>>>  	ret = set_sregs_arch206(vcpu, sregs);
>>>  	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>> +
>>> +	ret = vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>>>  
>>> -	return vcpu->kvm->arch.kvm_ops->set_sregs(vcpu, sregs);
>>> +out:
>>> +	vcpu_put(vcpu);
>>> +	return ret;
>>>  }
>>>  
>>>  int kvmppc_get_one_reg(struct kvm_vcpu *vcpu, u64 id,
>>> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
>>> index 18011fc4ac49..d95b4f15e52b 100644
>>> --- a/arch/s390/kvm/kvm-s390.c
>>> +++ b/arch/s390/kvm/kvm-s390.c
>>> @@ -2729,8 +2729,12 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs)
>>>  int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  				  struct kvm_sregs *sregs)
>>>  {
>>> +	vcpu_load(vcpu);
>>> +
>>>  	memcpy(&vcpu->run->s.regs.acrs, &sregs->acrs, sizeof(sregs->acrs));
>>>  	memcpy(&vcpu->arch.sie_block->gcr, &sregs->crs, sizeof(sregs->crs));
>>> +
>>> +	vcpu_put(vcpu);
>>>  	return 0;
>>>  }
>>>  
>>> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
>>> index 20a5f6776eea..a31a80aee0b9 100644
>>> --- a/arch/x86/kvm/x86.c
>>> +++ b/arch/x86/kvm/x86.c
>>> @@ -7500,15 +7500,19 @@ int kvm_arch_vcpu_ioctl_set_sregs(struct kvm_vcpu *vcpu,
>>>  	int mmu_reset_needed = 0;
>>>  	int pending_vec, max_bits, idx;
>>>  	struct desc_ptr dt;
>>> +	int ret;
>>> +
>>> +	vcpu_load(vcpu);
>>>  
>>> +	ret = -EINVAL;
>>
>> dito
> 
> Sure.

I'm doing it when applying.

Paolo

>> Reviewed-by: David Hildenbrand <david@redhat.com>

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:29 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c |  8 +++++++-
>  arch/powerpc/kvm/booke.c  | 15 +++++++++++----
>  arch/s390/kvm/kvm-s390.c  |  4 ++++
>  arch/x86/kvm/x86.c        | 13 ++++++++++---
>  virt/kvm/kvm_main.c       |  2 --
>  5 files changed, 32 insertions(+), 10 deletions(-)

With David's suggestions included:

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:29 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c |  8 +++++++-
>  arch/powerpc/kvm/booke.c  | 15 +++++++++++----
>  arch/s390/kvm/kvm-s390.c  |  4 ++++
>  arch/x86/kvm/x86.c        | 13 ++++++++++---
>  virt/kvm/kvm_main.c       |  2 --
>  5 files changed, 32 insertions(+), 10 deletions(-)

With David's suggestions included:

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 07/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_sregs

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:29 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_sregs().
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/book3s.c |  8 +++++++-
>  arch/powerpc/kvm/booke.c  | 15 +++++++++++----
>  arch/s390/kvm/kvm-s390.c  |  4 ++++
>  arch/x86/kvm/x86.c        | 13 ++++++++++---
>  virt/kvm/kvm_main.c       |  2 --
>  5 files changed, 32 insertions(+), 10 deletions(-)

With David's suggestions included:

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 08/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 11 +++++++++--
 arch/x86/kvm/x86.c       |  3 +++
 virt/kvm/arm/arm.c       |  3 +++
 virt/kvm/kvm_main.c      |  2 --
 4 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index d95b4f15e52b..396fc3db6d63 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2836,9 +2836,16 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret;
+
+	vcpu_load(vcpu);
+
 	/* CHECK_STOP and LOAD are not supported yet */
-	return is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED :
-				       KVM_MP_STATE_OPERATING;
+	ret = is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED :
+				      KVM_MP_STATE_OPERATING;
+
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index a31a80aee0b9..9bf62c336aa5 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7440,6 +7440,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	vcpu_load(vcpu);
+
 	kvm_apic_accept_events(vcpu);
 	if (vcpu->arch.mp_state == KVM_MP_STATE_HALTED &&
 					vcpu->arch.pv.pv_unhalted)
@@ -7447,6 +7449,7 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 	else
 		mp_state->mp_state = vcpu->arch.mp_state;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 1f448b208686..a7171701df44 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -381,11 +381,14 @@ static void vcpu_power_off(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	vcpu_load(vcpu);
+
 	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 19cf2d11f80f..eac3c29697db 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2603,9 +2603,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 	case KVM_GET_MP_STATE: {
 		struct kvm_mp_state mp_state;
 
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_mpstate(vcpu, &mp_state);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 08/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 11 +++++++++--
 arch/x86/kvm/x86.c       |  3 +++
 virt/kvm/arm/arm.c       |  3 +++
 virt/kvm/kvm_main.c      |  2 --
 4 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index d95b4f15e52b..396fc3db6d63 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2836,9 +2836,16 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret;
+
+	vcpu_load(vcpu);
+
 	/* CHECK_STOP and LOAD are not supported yet */
-	return is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED :
-				       KVM_MP_STATE_OPERATING;
+	ret = is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED :
+				      KVM_MP_STATE_OPERATING;
+
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index a31a80aee0b9..9bf62c336aa5 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7440,6 +7440,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	vcpu_load(vcpu);
+
 	kvm_apic_accept_events(vcpu);
 	if (vcpu->arch.mp_state = KVM_MP_STATE_HALTED &&
 					vcpu->arch.pv.pv_unhalted)
@@ -7447,6 +7449,7 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 	else
 		mp_state->mp_state = vcpu->arch.mp_state;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 1f448b208686..a7171701df44 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -381,11 +381,14 @@ static void vcpu_power_off(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	vcpu_load(vcpu);
+
 	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 19cf2d11f80f..eac3c29697db 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2603,9 +2603,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 	case KVM_GET_MP_STATE: {
 		struct kvm_mp_state mp_state;
 
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_mpstate(vcpu, &mp_state);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 08/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 11 +++++++++--
 arch/x86/kvm/x86.c       |  3 +++
 virt/kvm/arm/arm.c       |  3 +++
 virt/kvm/kvm_main.c      |  2 --
 4 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index d95b4f15e52b..396fc3db6d63 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2836,9 +2836,16 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret;
+
+	vcpu_load(vcpu);
+
 	/* CHECK_STOP and LOAD are not supported yet */
-	return is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED :
-				       KVM_MP_STATE_OPERATING;
+	ret = is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED :
+				      KVM_MP_STATE_OPERATING;
+
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index a31a80aee0b9..9bf62c336aa5 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7440,6 +7440,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	vcpu_load(vcpu);
+
 	kvm_apic_accept_events(vcpu);
 	if (vcpu->arch.mp_state == KVM_MP_STATE_HALTED &&
 					vcpu->arch.pv.pv_unhalted)
@@ -7447,6 +7449,7 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 	else
 		mp_state->mp_state = vcpu->arch.mp_state;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 1f448b208686..a7171701df44 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -381,11 +381,14 @@ static void vcpu_power_off(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	vcpu_load(vcpu);
+
 	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 19cf2d11f80f..eac3c29697db 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2603,9 +2603,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 	case KVM_GET_MP_STATE: {
 		struct kvm_mp_state mp_state;
 
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_mpstate(vcpu, &mp_state);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 08/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 11 +++++++++--
 arch/x86/kvm/x86.c       |  3 +++
 virt/kvm/arm/arm.c       |  3 +++
 virt/kvm/kvm_main.c      |  2 --
 4 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index d95b4f15e52b..396fc3db6d63 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2836,9 +2836,16 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret;
+
+	vcpu_load(vcpu);
+
 	/* CHECK_STOP and LOAD are not supported yet */
-	return is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED :
-				       KVM_MP_STATE_OPERATING;
+	ret = is_vcpu_stopped(vcpu) ? KVM_MP_STATE_STOPPED :
+				      KVM_MP_STATE_OPERATING;
+
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index a31a80aee0b9..9bf62c336aa5 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7440,6 +7440,8 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	vcpu_load(vcpu);
+
 	kvm_apic_accept_events(vcpu);
 	if (vcpu->arch.mp_state == KVM_MP_STATE_HALTED &&
 					vcpu->arch.pv.pv_unhalted)
@@ -7447,6 +7449,7 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 	else
 		mp_state->mp_state = vcpu->arch.mp_state;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 1f448b208686..a7171701df44 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -381,11 +381,14 @@ static void vcpu_power_off(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	vcpu_load(vcpu);
+
 	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 19cf2d11f80f..eac3c29697db 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2603,9 +2603,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 	case KVM_GET_MP_STATE: {
 		struct kvm_mp_state mp_state;
 
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_mpstate(vcpu, &mp_state);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

Re: [PATCH v3 08/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:30 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_mpstate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 11 +++++++++--
>  arch/x86/kvm/x86.c       |  3 +++
>  virt/kvm/arm/arm.c       |  3 +++
>  virt/kvm/kvm_main.c      |  2 --
>  4 files changed, 15 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 08/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:30 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_mpstate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 11 +++++++++--
>  arch/x86/kvm/x86.c       |  3 +++
>  virt/kvm/arm/arm.c       |  3 +++
>  virt/kvm/kvm_main.c      |  2 --
>  4 files changed, 15 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 08/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_mpstate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:30 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_mpstate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 11 +++++++++--
>  arch/x86/kvm/x86.c       |  3 +++
>  virt/kvm/arm/arm.c       |  3 +++
>  virt/kvm/kvm_main.c      |  2 --
>  4 files changed, 15 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 09/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c |  3 +++
 arch/x86/kvm/x86.c       | 14 +++++++++++---
 virt/kvm/arm/arm.c       |  9 +++++++--
 virt/kvm/kvm_main.c      |  2 --
 4 files changed, 21 insertions(+), 7 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 396fc3db6d63..8fade858c790 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2853,6 +2853,8 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	int rc = 0;
 
+	vcpu_load(vcpu);
+
 	/* user space knows about this interface - let it control the state */
 	vcpu->kvm->arch.user_cpu_state_ctrl = 1;
 
@@ -2870,6 +2872,7 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		rc = -ENXIO;
 	}
 
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9bf62c336aa5..8e67428af01b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7456,15 +7456,19 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret = -EINVAL;
+
+	vcpu_load(vcpu);
+
 	if (!lapic_in_kernel(vcpu) &&
 	    mp_state->mp_state != KVM_MP_STATE_RUNNABLE)
-		return -EINVAL;
+		goto out;
 
 	/* INITs are latched while in SMM */
 	if ((is_smm(vcpu) || vcpu->arch.smi_pending) &&
 	    (mp_state->mp_state == KVM_MP_STATE_SIPI_RECEIVED ||
 	     mp_state->mp_state == KVM_MP_STATE_INIT_RECEIVED))
-		return -EINVAL;
+		goto out;
 
 	if (mp_state->mp_state == KVM_MP_STATE_SIPI_RECEIVED) {
 		vcpu->arch.mp_state = KVM_MP_STATE_INIT_RECEIVED;
@@ -7472,7 +7476,11 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 	} else
 		vcpu->arch.mp_state = mp_state->mp_state;
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
-	return 0;
+
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int idt_index,
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a7171701df44..9a3acbcf542c 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -395,6 +395,10 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
 		vcpu->arch.power_off = false;
@@ -403,10 +407,11 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		vcpu_power_off(vcpu);
 		break;
 	default:
-		return -EINVAL;
+		ret = -EINVAL;
 	}
 
-	return 0;
+	vcpu_put(vcpu);
+	return ret;
 }
 
 /**
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index eac3c29697db..f3600052adbb 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2618,9 +2618,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&mp_state, argp, sizeof(mp_state)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_mpstate(vcpu, &mp_state);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_TRANSLATE: {
-- 
2.14.2

[PATCH v3 09/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c |  3 +++
 arch/x86/kvm/x86.c       | 14 +++++++++++---
 virt/kvm/arm/arm.c       |  9 +++++++--
 virt/kvm/kvm_main.c      |  2 --
 4 files changed, 21 insertions(+), 7 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 396fc3db6d63..8fade858c790 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2853,6 +2853,8 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	int rc = 0;
 
+	vcpu_load(vcpu);
+
 	/* user space knows about this interface - let it control the state */
 	vcpu->kvm->arch.user_cpu_state_ctrl = 1;
 
@@ -2870,6 +2872,7 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		rc = -ENXIO;
 	}
 
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9bf62c336aa5..8e67428af01b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7456,15 +7456,19 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret = -EINVAL;
+
+	vcpu_load(vcpu);
+
 	if (!lapic_in_kernel(vcpu) &&
 	    mp_state->mp_state != KVM_MP_STATE_RUNNABLE)
-		return -EINVAL;
+		goto out;
 
 	/* INITs are latched while in SMM */
 	if ((is_smm(vcpu) || vcpu->arch.smi_pending) &&
 	    (mp_state->mp_state = KVM_MP_STATE_SIPI_RECEIVED ||
 	     mp_state->mp_state = KVM_MP_STATE_INIT_RECEIVED))
-		return -EINVAL;
+		goto out;
 
 	if (mp_state->mp_state = KVM_MP_STATE_SIPI_RECEIVED) {
 		vcpu->arch.mp_state = KVM_MP_STATE_INIT_RECEIVED;
@@ -7472,7 +7476,11 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 	} else
 		vcpu->arch.mp_state = mp_state->mp_state;
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
-	return 0;
+
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int idt_index,
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a7171701df44..9a3acbcf542c 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -395,6 +395,10 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
 		vcpu->arch.power_off = false;
@@ -403,10 +407,11 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		vcpu_power_off(vcpu);
 		break;
 	default:
-		return -EINVAL;
+		ret = -EINVAL;
 	}
 
-	return 0;
+	vcpu_put(vcpu);
+	return ret;
 }
 
 /**
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index eac3c29697db..f3600052adbb 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2618,9 +2618,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&mp_state, argp, sizeof(mp_state)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_mpstate(vcpu, &mp_state);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_TRANSLATE: {
-- 
2.14.2

[PATCH v3 09/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c |  3 +++
 arch/x86/kvm/x86.c       | 14 +++++++++++---
 virt/kvm/arm/arm.c       |  9 +++++++--
 virt/kvm/kvm_main.c      |  2 --
 4 files changed, 21 insertions(+), 7 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 396fc3db6d63..8fade858c790 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2853,6 +2853,8 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	int rc = 0;
 
+	vcpu_load(vcpu);
+
 	/* user space knows about this interface - let it control the state */
 	vcpu->kvm->arch.user_cpu_state_ctrl = 1;
 
@@ -2870,6 +2872,7 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		rc = -ENXIO;
 	}
 
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9bf62c336aa5..8e67428af01b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7456,15 +7456,19 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret = -EINVAL;
+
+	vcpu_load(vcpu);
+
 	if (!lapic_in_kernel(vcpu) &&
 	    mp_state->mp_state != KVM_MP_STATE_RUNNABLE)
-		return -EINVAL;
+		goto out;
 
 	/* INITs are latched while in SMM */
 	if ((is_smm(vcpu) || vcpu->arch.smi_pending) &&
 	    (mp_state->mp_state == KVM_MP_STATE_SIPI_RECEIVED ||
 	     mp_state->mp_state == KVM_MP_STATE_INIT_RECEIVED))
-		return -EINVAL;
+		goto out;
 
 	if (mp_state->mp_state == KVM_MP_STATE_SIPI_RECEIVED) {
 		vcpu->arch.mp_state = KVM_MP_STATE_INIT_RECEIVED;
@@ -7472,7 +7476,11 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 	} else
 		vcpu->arch.mp_state = mp_state->mp_state;
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
-	return 0;
+
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int idt_index,
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a7171701df44..9a3acbcf542c 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -395,6 +395,10 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
 		vcpu->arch.power_off = false;
@@ -403,10 +407,11 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		vcpu_power_off(vcpu);
 		break;
 	default:
-		return -EINVAL;
+		ret = -EINVAL;
 	}
 
-	return 0;
+	vcpu_put(vcpu);
+	return ret;
 }
 
 /**
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index eac3c29697db..f3600052adbb 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2618,9 +2618,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&mp_state, argp, sizeof(mp_state)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_mpstate(vcpu, &mp_state);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_TRANSLATE: {
-- 
2.14.2

[PATCH v3 09/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_mpstate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c |  3 +++
 arch/x86/kvm/x86.c       | 14 +++++++++++---
 virt/kvm/arm/arm.c       |  9 +++++++--
 virt/kvm/kvm_main.c      |  2 --
 4 files changed, 21 insertions(+), 7 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 396fc3db6d63..8fade858c790 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2853,6 +2853,8 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	int rc = 0;
 
+	vcpu_load(vcpu);
+
 	/* user space knows about this interface - let it control the state */
 	vcpu->kvm->arch.user_cpu_state_ctrl = 1;
 
@@ -2870,6 +2872,7 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		rc = -ENXIO;
 	}
 
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 9bf62c336aa5..8e67428af01b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7456,15 +7456,19 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret = -EINVAL;
+
+	vcpu_load(vcpu);
+
 	if (!lapic_in_kernel(vcpu) &&
 	    mp_state->mp_state != KVM_MP_STATE_RUNNABLE)
-		return -EINVAL;
+		goto out;
 
 	/* INITs are latched while in SMM */
 	if ((is_smm(vcpu) || vcpu->arch.smi_pending) &&
 	    (mp_state->mp_state == KVM_MP_STATE_SIPI_RECEIVED ||
 	     mp_state->mp_state == KVM_MP_STATE_INIT_RECEIVED))
-		return -EINVAL;
+		goto out;
 
 	if (mp_state->mp_state == KVM_MP_STATE_SIPI_RECEIVED) {
 		vcpu->arch.mp_state = KVM_MP_STATE_INIT_RECEIVED;
@@ -7472,7 +7476,11 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 	} else
 		vcpu->arch.mp_state = mp_state->mp_state;
 	kvm_make_request(KVM_REQ_EVENT, vcpu);
-	return 0;
+
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int idt_index,
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a7171701df44..9a3acbcf542c 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -395,6 +395,10 @@ int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
 		vcpu->arch.power_off = false;
@@ -403,10 +407,11 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		vcpu_power_off(vcpu);
 		break;
 	default:
-		return -EINVAL;
+		ret = -EINVAL;
 	}
 
-	return 0;
+	vcpu_put(vcpu);
+	return ret;
 }
 
 /**
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index eac3c29697db..f3600052adbb 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2618,9 +2618,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&mp_state, argp, sizeof(mp_state)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_mpstate(vcpu, &mp_state);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_TRANSLATE: {
-- 
2.14.2

Re: [PATCH v3 09/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:31 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_mpstate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c |  3 +++
>  arch/x86/kvm/x86.c       | 14 +++++++++++---
>  virt/kvm/arm/arm.c       |  9 +++++++--
>  virt/kvm/kvm_main.c      |  2 --
>  4 files changed, 21 insertions(+), 7 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 09/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:31 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_mpstate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c |  3 +++
>  arch/x86/kvm/x86.c       | 14 +++++++++++---
>  virt/kvm/arm/arm.c       |  9 +++++++--
>  virt/kvm/kvm_main.c      |  2 --
>  4 files changed, 21 insertions(+), 7 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 09/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_mpstate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:31 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_mpstate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c |  3 +++
>  arch/x86/kvm/x86.c       | 14 +++++++++++---
>  virt/kvm/arm/arm.c       |  9 +++++++--
>  virt/kvm/kvm_main.c      |  2 --
>  4 files changed, 21 insertions(+), 7 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 10/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_translate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/booke.c | 2 ++
 arch/x86/kvm/x86.c       | 3 +++
 virt/kvm/kvm_main.c      | 2 --
 3 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index cdf0be02c95a..1b491b89cd43 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1793,7 +1793,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 {
 	int r;
 
+	vcpu_load(vcpu);
 	r = kvmppc_core_vcpu_translate(vcpu, tr);
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 8e67428af01b..c30ba99e7aa3 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7660,6 +7660,8 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	gpa_t gpa;
 	int idx;
 
+	vcpu_load(vcpu);
+
 	idx = srcu_read_lock(&vcpu->kvm->srcu);
 	gpa = kvm_mmu_gva_to_gpa_system(vcpu, vaddr, NULL);
 	srcu_read_unlock(&vcpu->kvm->srcu, idx);
@@ -7668,6 +7670,7 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	tr->writeable = 1;
 	tr->usermode = 0;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f3600052adbb..0a8a49073a23 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2627,9 +2627,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&tr, argp, sizeof(tr)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_translate(vcpu, &tr);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 10/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_translate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/booke.c | 2 ++
 arch/x86/kvm/x86.c       | 3 +++
 virt/kvm/kvm_main.c      | 2 --
 3 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index cdf0be02c95a..1b491b89cd43 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1793,7 +1793,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 {
 	int r;
 
+	vcpu_load(vcpu);
 	r = kvmppc_core_vcpu_translate(vcpu, tr);
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 8e67428af01b..c30ba99e7aa3 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7660,6 +7660,8 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	gpa_t gpa;
 	int idx;
 
+	vcpu_load(vcpu);
+
 	idx = srcu_read_lock(&vcpu->kvm->srcu);
 	gpa = kvm_mmu_gva_to_gpa_system(vcpu, vaddr, NULL);
 	srcu_read_unlock(&vcpu->kvm->srcu, idx);
@@ -7668,6 +7670,7 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	tr->writeable = 1;
 	tr->usermode = 0;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f3600052adbb..0a8a49073a23 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2627,9 +2627,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&tr, argp, sizeof(tr)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_translate(vcpu, &tr);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 10/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_translate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/booke.c | 2 ++
 arch/x86/kvm/x86.c       | 3 +++
 virt/kvm/kvm_main.c      | 2 --
 3 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index cdf0be02c95a..1b491b89cd43 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1793,7 +1793,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 {
 	int r;
 
+	vcpu_load(vcpu);
 	r = kvmppc_core_vcpu_translate(vcpu, tr);
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 8e67428af01b..c30ba99e7aa3 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7660,6 +7660,8 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	gpa_t gpa;
 	int idx;
 
+	vcpu_load(vcpu);
+
 	idx = srcu_read_lock(&vcpu->kvm->srcu);
 	gpa = kvm_mmu_gva_to_gpa_system(vcpu, vaddr, NULL);
 	srcu_read_unlock(&vcpu->kvm->srcu, idx);
@@ -7668,6 +7670,7 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	tr->writeable = 1;
 	tr->usermode = 0;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f3600052adbb..0a8a49073a23 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2627,9 +2627,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&tr, argp, sizeof(tr)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_translate(vcpu, &tr);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 10/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_translate().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/powerpc/kvm/booke.c | 2 ++
 arch/x86/kvm/x86.c       | 3 +++
 virt/kvm/kvm_main.c      | 2 --
 3 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index cdf0be02c95a..1b491b89cd43 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -1793,7 +1793,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 {
 	int r;
 
+	vcpu_load(vcpu);
 	r = kvmppc_core_vcpu_translate(vcpu, tr);
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 8e67428af01b..c30ba99e7aa3 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7660,6 +7660,8 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	gpa_t gpa;
 	int idx;
 
+	vcpu_load(vcpu);
+
 	idx = srcu_read_lock(&vcpu->kvm->srcu);
 	gpa = kvm_mmu_gva_to_gpa_system(vcpu, vaddr, NULL);
 	srcu_read_unlock(&vcpu->kvm->srcu, idx);
@@ -7668,6 +7670,7 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 	tr->writeable = 1;
 	tr->usermode = 0;
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index f3600052adbb..0a8a49073a23 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2627,9 +2627,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&tr, argp, sizeof(tr)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_translate(vcpu, &tr);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

Re: [PATCH v3 10/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:32 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_translate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/booke.c | 2 ++
>  arch/x86/kvm/x86.c       | 3 +++
>  virt/kvm/kvm_main.c      | 2 --
>  3 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 10/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:32 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_translate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/booke.c | 2 ++
>  arch/x86/kvm/x86.c       | 3 +++
>  virt/kvm/kvm_main.c      | 2 --
>  3 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 10/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_translate

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:32 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_translate().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/powerpc/kvm/booke.c | 2 ++
>  arch/x86/kvm/x86.c       | 3 +++
>  virt/kvm/kvm_main.c      | 2 --
>  3 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_guest_debug().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm64/kvm/guest.c    | 15 ++++++++++++---
 arch/powerpc/kvm/book3s.c |  2 ++
 arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
 arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
 arch/x86/kvm/x86.c        |  4 +++-
 virt/kvm/kvm_main.c       |  2 --
 6 files changed, 42 insertions(+), 16 deletions(-)

diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index 5c7f657dd207..d7e3299a7734 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -361,10 +361,16 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
 	trace_kvm_set_guest_debug(vcpu, dbg->control);
 
-	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK)
-		return -EINVAL;
+	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK) {
+		ret = -EINVAL;
+		goto out;
+	}
 
 	if (dbg->control & KVM_GUESTDBG_ENABLE) {
 		vcpu->guest_debug = dbg->control;
@@ -378,7 +384,10 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		/* If not enabled clear all flags */
 		vcpu->guest_debug = 0;
 	}
-	return 0;
+
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arm_vcpu_arch_set_attr(struct kvm_vcpu *vcpu,
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 047651622cb8..234531d1bee1 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -755,7 +755,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
+	vcpu_load(vcpu);
 	vcpu->guest_debug = dbg->control;
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 1b491b89cd43..7cb0e2677e60 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	struct debug_reg *dbg_reg;
 	int n, b = 0, w = 0;
+	int ret = 0;
+
+	vcpu_load(vcpu);
 
 	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
 		vcpu->arch.dbg_reg.dbcr0 = 0;
 		vcpu->guest_debug = 0;
 		kvm_guest_protect_msr(vcpu, MSR_DE, false);
-		return 0;
+		goto out;
 	}
 
 	kvm_guest_protect_msr(vcpu, MSR_DE, true);
@@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 #endif
 
 	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
-		return 0;
+		goto out;
 
+	ret = -EINVAL;
 	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
 		uint64_t addr = dbg->arch.bp[n].addr;
 		uint32_t type = dbg->arch.bp[n].type;
@@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
 			     KVMPPC_DEBUG_WATCH_WRITE |
 			     KVMPPC_DEBUG_BREAKPOINT))
-			return -EINVAL;
+			goto out;
 
 		if (type & KVMPPC_DEBUG_BREAKPOINT) {
 			/* Setting H/W breakpoint */
 			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
-				return -EINVAL;
+				goto out;
 		} else {
 			/* Setting H/W watchpoint */
 			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
 							type, w++))
-				return -EINVAL;
+				goto out;
 		}
 	}
 
-	return 0;
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 8fade858c790..4bf80b57b5c1 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2804,13 +2804,19 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	int rc = 0;
 
+	vcpu_load(vcpu);
+
 	vcpu->guest_debug = 0;
 	kvm_s390_clear_bp_data(vcpu);
 
-	if (dbg->control & ~VALID_GUESTDBG_FLAGS)
-		return -EINVAL;
-	if (!sclp.has_gpere)
-		return -EINVAL;
+	if (dbg->control & ~VALID_GUESTDBG_FLAGS) {
+		rc = -EINVAL;
+		goto out;
+	}
+	if (!sclp.has_gpere) {
+		rc = -EINVAL;
+		goto out;
+	}
 
 	if (dbg->control & KVM_GUESTDBG_ENABLE) {
 		vcpu->guest_debug = dbg->control;
@@ -2830,6 +2836,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		atomic_andnot(CPUSTAT_P, &vcpu->arch.sie_block->cpuflags);
 	}
 
+out:
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index c30ba99e7aa3..5d19caee6d51 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7601,6 +7601,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	unsigned long rflags;
 	int i, r;
 
+	vcpu_load(vcpu);
+
 	if (dbg->control & (KVM_GUESTDBG_INJECT_DB | KVM_GUESTDBG_INJECT_BP)) {
 		r = -EBUSY;
 		if (vcpu->arch.exception.pending)
@@ -7646,7 +7648,7 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	r = 0;
 
 out:
-
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 0a8a49073a23..c688eb777bec 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2642,9 +2642,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&dbg, argp, sizeof(dbg)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_guest_debug(vcpu, &dbg);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_SET_SIGNAL_MASK: {
-- 
2.14.2

[PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_guest_debug().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm64/kvm/guest.c    | 15 ++++++++++++---
 arch/powerpc/kvm/book3s.c |  2 ++
 arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
 arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
 arch/x86/kvm/x86.c        |  4 +++-
 virt/kvm/kvm_main.c       |  2 --
 6 files changed, 42 insertions(+), 16 deletions(-)

diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index 5c7f657dd207..d7e3299a7734 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -361,10 +361,16 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
 	trace_kvm_set_guest_debug(vcpu, dbg->control);
 
-	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK)
-		return -EINVAL;
+	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK) {
+		ret = -EINVAL;
+		goto out;
+	}
 
 	if (dbg->control & KVM_GUESTDBG_ENABLE) {
 		vcpu->guest_debug = dbg->control;
@@ -378,7 +384,10 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		/* If not enabled clear all flags */
 		vcpu->guest_debug = 0;
 	}
-	return 0;
+
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arm_vcpu_arch_set_attr(struct kvm_vcpu *vcpu,
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 047651622cb8..234531d1bee1 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -755,7 +755,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
+	vcpu_load(vcpu);
 	vcpu->guest_debug = dbg->control;
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 1b491b89cd43..7cb0e2677e60 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	struct debug_reg *dbg_reg;
 	int n, b = 0, w = 0;
+	int ret = 0;
+
+	vcpu_load(vcpu);
 
 	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
 		vcpu->arch.dbg_reg.dbcr0 = 0;
 		vcpu->guest_debug = 0;
 		kvm_guest_protect_msr(vcpu, MSR_DE, false);
-		return 0;
+		goto out;
 	}
 
 	kvm_guest_protect_msr(vcpu, MSR_DE, true);
@@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 #endif
 
 	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
-		return 0;
+		goto out;
 
+	ret = -EINVAL;
 	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
 		uint64_t addr = dbg->arch.bp[n].addr;
 		uint32_t type = dbg->arch.bp[n].type;
@@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
 			     KVMPPC_DEBUG_WATCH_WRITE |
 			     KVMPPC_DEBUG_BREAKPOINT))
-			return -EINVAL;
+			goto out;
 
 		if (type & KVMPPC_DEBUG_BREAKPOINT) {
 			/* Setting H/W breakpoint */
 			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
-				return -EINVAL;
+				goto out;
 		} else {
 			/* Setting H/W watchpoint */
 			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
 							type, w++))
-				return -EINVAL;
+				goto out;
 		}
 	}
 
-	return 0;
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 8fade858c790..4bf80b57b5c1 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2804,13 +2804,19 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	int rc = 0;
 
+	vcpu_load(vcpu);
+
 	vcpu->guest_debug = 0;
 	kvm_s390_clear_bp_data(vcpu);
 
-	if (dbg->control & ~VALID_GUESTDBG_FLAGS)
-		return -EINVAL;
-	if (!sclp.has_gpere)
-		return -EINVAL;
+	if (dbg->control & ~VALID_GUESTDBG_FLAGS) {
+		rc = -EINVAL;
+		goto out;
+	}
+	if (!sclp.has_gpere) {
+		rc = -EINVAL;
+		goto out;
+	}
 
 	if (dbg->control & KVM_GUESTDBG_ENABLE) {
 		vcpu->guest_debug = dbg->control;
@@ -2830,6 +2836,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		atomic_andnot(CPUSTAT_P, &vcpu->arch.sie_block->cpuflags);
 	}
 
+out:
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index c30ba99e7aa3..5d19caee6d51 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7601,6 +7601,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	unsigned long rflags;
 	int i, r;
 
+	vcpu_load(vcpu);
+
 	if (dbg->control & (KVM_GUESTDBG_INJECT_DB | KVM_GUESTDBG_INJECT_BP)) {
 		r = -EBUSY;
 		if (vcpu->arch.exception.pending)
@@ -7646,7 +7648,7 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	r = 0;
 
 out:
-
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 0a8a49073a23..c688eb777bec 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2642,9 +2642,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&dbg, argp, sizeof(dbg)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_guest_debug(vcpu, &dbg);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_SET_SIGNAL_MASK: {
-- 
2.14.2

[PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_guest_debug().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm64/kvm/guest.c    | 15 ++++++++++++---
 arch/powerpc/kvm/book3s.c |  2 ++
 arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
 arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
 arch/x86/kvm/x86.c        |  4 +++-
 virt/kvm/kvm_main.c       |  2 --
 6 files changed, 42 insertions(+), 16 deletions(-)

diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index 5c7f657dd207..d7e3299a7734 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -361,10 +361,16 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
 	trace_kvm_set_guest_debug(vcpu, dbg->control);
 
-	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK)
-		return -EINVAL;
+	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK) {
+		ret = -EINVAL;
+		goto out;
+	}
 
 	if (dbg->control & KVM_GUESTDBG_ENABLE) {
 		vcpu->guest_debug = dbg->control;
@@ -378,7 +384,10 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		/* If not enabled clear all flags */
 		vcpu->guest_debug = 0;
 	}
-	return 0;
+
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arm_vcpu_arch_set_attr(struct kvm_vcpu *vcpu,
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 047651622cb8..234531d1bee1 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -755,7 +755,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
+	vcpu_load(vcpu);
 	vcpu->guest_debug = dbg->control;
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 1b491b89cd43..7cb0e2677e60 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	struct debug_reg *dbg_reg;
 	int n, b = 0, w = 0;
+	int ret = 0;
+
+	vcpu_load(vcpu);
 
 	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
 		vcpu->arch.dbg_reg.dbcr0 = 0;
 		vcpu->guest_debug = 0;
 		kvm_guest_protect_msr(vcpu, MSR_DE, false);
-		return 0;
+		goto out;
 	}
 
 	kvm_guest_protect_msr(vcpu, MSR_DE, true);
@@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 #endif
 
 	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
-		return 0;
+		goto out;
 
+	ret = -EINVAL;
 	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
 		uint64_t addr = dbg->arch.bp[n].addr;
 		uint32_t type = dbg->arch.bp[n].type;
@@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
 			     KVMPPC_DEBUG_WATCH_WRITE |
 			     KVMPPC_DEBUG_BREAKPOINT))
-			return -EINVAL;
+			goto out;
 
 		if (type & KVMPPC_DEBUG_BREAKPOINT) {
 			/* Setting H/W breakpoint */
 			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
-				return -EINVAL;
+				goto out;
 		} else {
 			/* Setting H/W watchpoint */
 			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
 							type, w++))
-				return -EINVAL;
+				goto out;
 		}
 	}
 
-	return 0;
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 8fade858c790..4bf80b57b5c1 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2804,13 +2804,19 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	int rc = 0;
 
+	vcpu_load(vcpu);
+
 	vcpu->guest_debug = 0;
 	kvm_s390_clear_bp_data(vcpu);
 
-	if (dbg->control & ~VALID_GUESTDBG_FLAGS)
-		return -EINVAL;
-	if (!sclp.has_gpere)
-		return -EINVAL;
+	if (dbg->control & ~VALID_GUESTDBG_FLAGS) {
+		rc = -EINVAL;
+		goto out;
+	}
+	if (!sclp.has_gpere) {
+		rc = -EINVAL;
+		goto out;
+	}
 
 	if (dbg->control & KVM_GUESTDBG_ENABLE) {
 		vcpu->guest_debug = dbg->control;
@@ -2830,6 +2836,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		atomic_andnot(CPUSTAT_P, &vcpu->arch.sie_block->cpuflags);
 	}
 
+out:
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index c30ba99e7aa3..5d19caee6d51 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7601,6 +7601,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	unsigned long rflags;
 	int i, r;
 
+	vcpu_load(vcpu);
+
 	if (dbg->control & (KVM_GUESTDBG_INJECT_DB | KVM_GUESTDBG_INJECT_BP)) {
 		r = -EBUSY;
 		if (vcpu->arch.exception.pending)
@@ -7646,7 +7648,7 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	r = 0;
 
 out:
-
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 0a8a49073a23..c688eb777bec 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2642,9 +2642,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&dbg, argp, sizeof(dbg)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_guest_debug(vcpu, &dbg);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_SET_SIGNAL_MASK: {
-- 
2.14.2

[PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_guest_debug().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm64/kvm/guest.c    | 15 ++++++++++++---
 arch/powerpc/kvm/book3s.c |  2 ++
 arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
 arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
 arch/x86/kvm/x86.c        |  4 +++-
 virt/kvm/kvm_main.c       |  2 --
 6 files changed, 42 insertions(+), 16 deletions(-)

diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index 5c7f657dd207..d7e3299a7734 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -361,10 +361,16 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
 	trace_kvm_set_guest_debug(vcpu, dbg->control);
 
-	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK)
-		return -EINVAL;
+	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK) {
+		ret = -EINVAL;
+		goto out;
+	}
 
 	if (dbg->control & KVM_GUESTDBG_ENABLE) {
 		vcpu->guest_debug = dbg->control;
@@ -378,7 +384,10 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		/* If not enabled clear all flags */
 		vcpu->guest_debug = 0;
 	}
-	return 0;
+
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arm_vcpu_arch_set_attr(struct kvm_vcpu *vcpu,
diff --git a/arch/powerpc/kvm/book3s.c b/arch/powerpc/kvm/book3s.c
index 047651622cb8..234531d1bee1 100644
--- a/arch/powerpc/kvm/book3s.c
+++ b/arch/powerpc/kvm/book3s.c
@@ -755,7 +755,9 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 					struct kvm_guest_debug *dbg)
 {
+	vcpu_load(vcpu);
 	vcpu->guest_debug = dbg->control;
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
index 1b491b89cd43..7cb0e2677e60 100644
--- a/arch/powerpc/kvm/booke.c
+++ b/arch/powerpc/kvm/booke.c
@@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	struct debug_reg *dbg_reg;
 	int n, b = 0, w = 0;
+	int ret = 0;
+
+	vcpu_load(vcpu);
 
 	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
 		vcpu->arch.dbg_reg.dbcr0 = 0;
 		vcpu->guest_debug = 0;
 		kvm_guest_protect_msr(vcpu, MSR_DE, false);
-		return 0;
+		goto out;
 	}
 
 	kvm_guest_protect_msr(vcpu, MSR_DE, true);
@@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 #endif
 
 	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
-		return 0;
+		goto out;
 
+	ret = -EINVAL;
 	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
 		uint64_t addr = dbg->arch.bp[n].addr;
 		uint32_t type = dbg->arch.bp[n].type;
@@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
 			     KVMPPC_DEBUG_WATCH_WRITE |
 			     KVMPPC_DEBUG_BREAKPOINT))
-			return -EINVAL;
+			goto out;
 
 		if (type & KVMPPC_DEBUG_BREAKPOINT) {
 			/* Setting H/W breakpoint */
 			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
-				return -EINVAL;
+				goto out;
 		} else {
 			/* Setting H/W watchpoint */
 			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
 							type, w++))
-				return -EINVAL;
+				goto out;
 		}
 	}
 
-	return 0;
+	ret = 0;
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 8fade858c790..4bf80b57b5c1 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2804,13 +2804,19 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	int rc = 0;
 
+	vcpu_load(vcpu);
+
 	vcpu->guest_debug = 0;
 	kvm_s390_clear_bp_data(vcpu);
 
-	if (dbg->control & ~VALID_GUESTDBG_FLAGS)
-		return -EINVAL;
-	if (!sclp.has_gpere)
-		return -EINVAL;
+	if (dbg->control & ~VALID_GUESTDBG_FLAGS) {
+		rc = -EINVAL;
+		goto out;
+	}
+	if (!sclp.has_gpere) {
+		rc = -EINVAL;
+		goto out;
+	}
 
 	if (dbg->control & KVM_GUESTDBG_ENABLE) {
 		vcpu->guest_debug = dbg->control;
@@ -2830,6 +2836,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 		atomic_andnot(CPUSTAT_P, &vcpu->arch.sie_block->cpuflags);
 	}
 
+out:
+	vcpu_put(vcpu);
 	return rc;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index c30ba99e7aa3..5d19caee6d51 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7601,6 +7601,8 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	unsigned long rflags;
 	int i, r;
 
+	vcpu_load(vcpu);
+
 	if (dbg->control & (KVM_GUESTDBG_INJECT_DB | KVM_GUESTDBG_INJECT_BP)) {
 		r = -EBUSY;
 		if (vcpu->arch.exception.pending)
@@ -7646,7 +7648,7 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	r = 0;
 
 out:
-
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 0a8a49073a23..c688eb777bec 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2642,9 +2642,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -EFAULT;
 		if (copy_from_user(&dbg, argp, sizeof(dbg)))
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_guest_debug(vcpu, &dbg);
-		vcpu_put(vcpu);
 		break;
 	}
 	case KVM_SET_SIGNAL_MASK: {
-- 
2.14.2

Re: [PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:33 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_guest_debug().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/arm64/kvm/guest.c    | 15 ++++++++++++---
>  arch/powerpc/kvm/book3s.c |  2 ++
>  arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
>  arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
>  arch/x86/kvm/x86.c        |  4 +++-
>  virt/kvm/kvm_main.c       |  2 --
>  6 files changed, 42 insertions(+), 16 deletions(-)
> 

> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index 1b491b89cd43..7cb0e2677e60 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  {
>  	struct debug_reg *dbg_reg;
>  	int n, b = 0, w = 0;
> +	int ret = 0;
> +
> +	vcpu_load(vcpu);
>  
>  	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
>  		vcpu->arch.dbg_reg.dbcr0 = 0;
>  		vcpu->guest_debug = 0;
>  		kvm_guest_protect_msr(vcpu, MSR_DE, false);
> -		return 0;
> +		goto out;
>  	}
>  
>  	kvm_guest_protect_msr(vcpu, MSR_DE, true);
> @@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  #endif
>  
>  	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
> -		return 0;
> +		goto out;
>  
> +	ret = -EINVAL;
>  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
>  		uint64_t addr = dbg->arch.bp[n].addr;
>  		uint32_t type = dbg->arch.bp[n].type;
> @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
>  			     KVMPPC_DEBUG_WATCH_WRITE |
>  			     KVMPPC_DEBUG_BREAKPOINT))
> -			return -EINVAL;
> +			goto out;
>  
>  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
>  			/* Setting H/W breakpoint */
>  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
> -				return -EINVAL;
> +				goto out;
>  		} else {
>  			/* Setting H/W watchpoint */
>  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
>  							type, w++))
> -				return -EINVAL;
> +				goto out;
>  		}
>  	}
>  
> -	return 0;
> +	ret = 0;

I would probably set the -EINVAL in the individual branches (so it is
clear that something is wrong, and it is not just a benign exit as in
the cases above), but your code is correct as well. Let the powerpc
folks decide.

> +out:
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)

In any case,

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:33 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_guest_debug().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/arm64/kvm/guest.c    | 15 ++++++++++++---
>  arch/powerpc/kvm/book3s.c |  2 ++
>  arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
>  arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
>  arch/x86/kvm/x86.c        |  4 +++-
>  virt/kvm/kvm_main.c       |  2 --
>  6 files changed, 42 insertions(+), 16 deletions(-)
> 

> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index 1b491b89cd43..7cb0e2677e60 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  {
>  	struct debug_reg *dbg_reg;
>  	int n, b = 0, w = 0;
> +	int ret = 0;
> +
> +	vcpu_load(vcpu);
>  
>  	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
>  		vcpu->arch.dbg_reg.dbcr0 = 0;
>  		vcpu->guest_debug = 0;
>  		kvm_guest_protect_msr(vcpu, MSR_DE, false);
> -		return 0;
> +		goto out;
>  	}
>  
>  	kvm_guest_protect_msr(vcpu, MSR_DE, true);
> @@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  #endif
>  
>  	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
> -		return 0;
> +		goto out;
>  
> +	ret = -EINVAL;
>  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
>  		uint64_t addr = dbg->arch.bp[n].addr;
>  		uint32_t type = dbg->arch.bp[n].type;
> @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
>  			     KVMPPC_DEBUG_WATCH_WRITE |
>  			     KVMPPC_DEBUG_BREAKPOINT))
> -			return -EINVAL;
> +			goto out;
>  
>  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
>  			/* Setting H/W breakpoint */
>  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
> -				return -EINVAL;
> +				goto out;
>  		} else {
>  			/* Setting H/W watchpoint */
>  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
>  							type, w++))
> -				return -EINVAL;
> +				goto out;
>  		}
>  	}
>  
> -	return 0;
> +	ret = 0;

I would probably set the -EINVAL in the individual branches (so it is
clear that something is wrong, and it is not just a benign exit as in
the cases above), but your code is correct as well. Let the powerpc
folks decide.

> +out:
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)

In any case,

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:33 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_guest_debug().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/arm64/kvm/guest.c    | 15 ++++++++++++---
>  arch/powerpc/kvm/book3s.c |  2 ++
>  arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
>  arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
>  arch/x86/kvm/x86.c        |  4 +++-
>  virt/kvm/kvm_main.c       |  2 --
>  6 files changed, 42 insertions(+), 16 deletions(-)
> 

> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index 1b491b89cd43..7cb0e2677e60 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  {
>  	struct debug_reg *dbg_reg;
>  	int n, b = 0, w = 0;
> +	int ret = 0;
> +
> +	vcpu_load(vcpu);
>  
>  	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
>  		vcpu->arch.dbg_reg.dbcr0 = 0;
>  		vcpu->guest_debug = 0;
>  		kvm_guest_protect_msr(vcpu, MSR_DE, false);
> -		return 0;
> +		goto out;
>  	}
>  
>  	kvm_guest_protect_msr(vcpu, MSR_DE, true);
> @@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  #endif
>  
>  	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
> -		return 0;
> +		goto out;
>  
> +	ret = -EINVAL;
>  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
>  		uint64_t addr = dbg->arch.bp[n].addr;
>  		uint32_t type = dbg->arch.bp[n].type;
> @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
>  			     KVMPPC_DEBUG_WATCH_WRITE |
>  			     KVMPPC_DEBUG_BREAKPOINT))
> -			return -EINVAL;
> +			goto out;
>  
>  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
>  			/* Setting H/W breakpoint */
>  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
> -				return -EINVAL;
> +				goto out;
>  		} else {
>  			/* Setting H/W watchpoint */
>  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
>  							type, w++))
> -				return -EINVAL;
> +				goto out;
>  		}
>  	}
>  
> -	return 0;
> +	ret = 0;

I would probably set the -EINVAL in the individual branches (so it is
clear that something is wrong, and it is not just a benign exit as in
the cases above), but your code is correct as well. Let the powerpc
folks decide.

> +out:
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)

In any case,

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:33 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_guest_debug().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/arm64/kvm/guest.c    | 15 ++++++++++++---
>  arch/powerpc/kvm/book3s.c |  2 ++
>  arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
>  arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
>  arch/x86/kvm/x86.c        |  4 +++-
>  virt/kvm/kvm_main.c       |  2 --
>  6 files changed, 42 insertions(+), 16 deletions(-)
> 

> diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> index 1b491b89cd43..7cb0e2677e60 100644
> --- a/arch/powerpc/kvm/booke.c
> +++ b/arch/powerpc/kvm/booke.c
> @@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  {
>  	struct debug_reg *dbg_reg;
>  	int n, b = 0, w = 0;
> +	int ret = 0;
> +
> +	vcpu_load(vcpu);
>  
>  	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
>  		vcpu->arch.dbg_reg.dbcr0 = 0;
>  		vcpu->guest_debug = 0;
>  		kvm_guest_protect_msr(vcpu, MSR_DE, false);
> -		return 0;
> +		goto out;
>  	}
>  
>  	kvm_guest_protect_msr(vcpu, MSR_DE, true);
> @@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  #endif
>  
>  	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
> -		return 0;
> +		goto out;
>  
> +	ret = -EINVAL;
>  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
>  		uint64_t addr = dbg->arch.bp[n].addr;
>  		uint32_t type = dbg->arch.bp[n].type;
> @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
>  			     KVMPPC_DEBUG_WATCH_WRITE |
>  			     KVMPPC_DEBUG_BREAKPOINT))
> -			return -EINVAL;
> +			goto out;
>  
>  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
>  			/* Setting H/W breakpoint */
>  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
> -				return -EINVAL;
> +				goto out;
>  		} else {
>  			/* Setting H/W watchpoint */
>  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
>  							type, w++))
> -				return -EINVAL;
> +				goto out;
>  		}
>  	}
>  
> -	return 0;
> +	ret = 0;

I would probably set the -EINVAL in the individual branches (so it is
clear that something is wrong, and it is not just a benign exit as in
the cases above), but your code is correct as well. Let the powerpc
folks decide.

> +out:
> +	vcpu_put(vcpu);
> +	return ret;
>  }
>  
>  void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)

In any case,

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Christoffer Dall]

On Mon, Dec 11, 2017 at 01:39:43PM +0100, Cornelia Huck wrote:
> On Mon,  4 Dec 2017 21:35:33 +0100
> Christoffer Dall <cdall@kernel.org> wrote:
> 
> > From: Christoffer Dall <christoffer.dall@linaro.org>
> > 
> > Move vcpu_load() and vcpu_put() into the architecture specific
> > implementations of kvm_arch_vcpu_ioctl_set_guest_debug().
> > 
> > Reviewed-by: David Hildenbrand <david@redhat.com>
> > Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> > ---
> >  arch/arm64/kvm/guest.c    | 15 ++++++++++++---
> >  arch/powerpc/kvm/book3s.c |  2 ++
> >  arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
> >  arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
> >  arch/x86/kvm/x86.c        |  4 +++-
> >  virt/kvm/kvm_main.c       |  2 --
> >  6 files changed, 42 insertions(+), 16 deletions(-)
> > 
> 
> > diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> > index 1b491b89cd43..7cb0e2677e60 100644
> > --- a/arch/powerpc/kvm/booke.c
> > +++ b/arch/powerpc/kvm/booke.c
> > @@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  {
> >  	struct debug_reg *dbg_reg;
> >  	int n, b = 0, w = 0;
> > +	int ret = 0;
> > +
> > +	vcpu_load(vcpu);
> >  
> >  	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
> >  		vcpu->arch.dbg_reg.dbcr0 = 0;
> >  		vcpu->guest_debug = 0;
> >  		kvm_guest_protect_msr(vcpu, MSR_DE, false);
> > -		return 0;
> > +		goto out;
> >  	}
> >  
> >  	kvm_guest_protect_msr(vcpu, MSR_DE, true);
> > @@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  #endif
> >  
> >  	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
> > -		return 0;
> > +		goto out;
> >  
> > +	ret = -EINVAL;
> >  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
> >  		uint64_t addr = dbg->arch.bp[n].addr;
> >  		uint32_t type = dbg->arch.bp[n].type;
> > @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
> >  			     KVMPPC_DEBUG_WATCH_WRITE |
> >  			     KVMPPC_DEBUG_BREAKPOINT))
> > -			return -EINVAL;
> > +			goto out;
> >  
> >  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
> >  			/* Setting H/W breakpoint */
> >  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
> > -				return -EINVAL;
> > +				goto out;
> >  		} else {
> >  			/* Setting H/W watchpoint */
> >  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
> >  							type, w++))
> > -				return -EINVAL;
> > +				goto out;
> >  		}
> >  	}
> >  
> > -	return 0;
> > +	ret = 0;
> 
> I would probably set the -EINVAL in the individual branches (so it is
> clear that something is wrong, and it is not just a benign exit as in
> the cases above), but your code is correct as well.

I think that's better as well actually.  I got probably got a little
used to that pattern after looking the main dispatcher function for a
while.  I'm happy to change it.

> > +out:
> > +	vcpu_put(vcpu);
> > +	return ret;
> >  }
> >  
> >  void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
> 
> In any case,
> 
> Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Thanks!
-Christoffer

Re: [PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Christoffer Dall]

On Mon, Dec 11, 2017 at 01:39:43PM +0100, Cornelia Huck wrote:
> On Mon,  4 Dec 2017 21:35:33 +0100
> Christoffer Dall <cdall@kernel.org> wrote:
> 
> > From: Christoffer Dall <christoffer.dall@linaro.org>
> > 
> > Move vcpu_load() and vcpu_put() into the architecture specific
> > implementations of kvm_arch_vcpu_ioctl_set_guest_debug().
> > 
> > Reviewed-by: David Hildenbrand <david@redhat.com>
> > Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> > ---
> >  arch/arm64/kvm/guest.c    | 15 ++++++++++++---
> >  arch/powerpc/kvm/book3s.c |  2 ++
> >  arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
> >  arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
> >  arch/x86/kvm/x86.c        |  4 +++-
> >  virt/kvm/kvm_main.c       |  2 --
> >  6 files changed, 42 insertions(+), 16 deletions(-)
> > 
> 
> > diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> > index 1b491b89cd43..7cb0e2677e60 100644
> > --- a/arch/powerpc/kvm/booke.c
> > +++ b/arch/powerpc/kvm/booke.c
> > @@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  {
> >  	struct debug_reg *dbg_reg;
> >  	int n, b = 0, w = 0;
> > +	int ret = 0;
> > +
> > +	vcpu_load(vcpu);
> >  
> >  	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
> >  		vcpu->arch.dbg_reg.dbcr0 = 0;
> >  		vcpu->guest_debug = 0;
> >  		kvm_guest_protect_msr(vcpu, MSR_DE, false);
> > -		return 0;
> > +		goto out;
> >  	}
> >  
> >  	kvm_guest_protect_msr(vcpu, MSR_DE, true);
> > @@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  #endif
> >  
> >  	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
> > -		return 0;
> > +		goto out;
> >  
> > +	ret = -EINVAL;
> >  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
> >  		uint64_t addr = dbg->arch.bp[n].addr;
> >  		uint32_t type = dbg->arch.bp[n].type;
> > @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
> >  			     KVMPPC_DEBUG_WATCH_WRITE |
> >  			     KVMPPC_DEBUG_BREAKPOINT))
> > -			return -EINVAL;
> > +			goto out;
> >  
> >  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
> >  			/* Setting H/W breakpoint */
> >  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
> > -				return -EINVAL;
> > +				goto out;
> >  		} else {
> >  			/* Setting H/W watchpoint */
> >  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
> >  							type, w++))
> > -				return -EINVAL;
> > +				goto out;
> >  		}
> >  	}
> >  
> > -	return 0;
> > +	ret = 0;
> 
> I would probably set the -EINVAL in the individual branches (so it is
> clear that something is wrong, and it is not just a benign exit as in
> the cases above), but your code is correct as well.

I think that's better as well actually.  I got probably got a little
used to that pattern after looking the main dispatcher function for a
while.  I'm happy to change it.

> > +out:
> > +	vcpu_put(vcpu);
> > +	return ret;
> >  }
> >  
> >  void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
> 
> In any case,
> 
> Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Thanks!
-Christoffer

[PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Christoffer Dall]

On Mon, Dec 11, 2017 at 01:39:43PM +0100, Cornelia Huck wrote:
> On Mon,  4 Dec 2017 21:35:33 +0100
> Christoffer Dall <cdall@kernel.org> wrote:
> 
> > From: Christoffer Dall <christoffer.dall@linaro.org>
> > 
> > Move vcpu_load() and vcpu_put() into the architecture specific
> > implementations of kvm_arch_vcpu_ioctl_set_guest_debug().
> > 
> > Reviewed-by: David Hildenbrand <david@redhat.com>
> > Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> > ---
> >  arch/arm64/kvm/guest.c    | 15 ++++++++++++---
> >  arch/powerpc/kvm/book3s.c |  2 ++
> >  arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
> >  arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
> >  arch/x86/kvm/x86.c        |  4 +++-
> >  virt/kvm/kvm_main.c       |  2 --
> >  6 files changed, 42 insertions(+), 16 deletions(-)
> > 
> 
> > diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> > index 1b491b89cd43..7cb0e2677e60 100644
> > --- a/arch/powerpc/kvm/booke.c
> > +++ b/arch/powerpc/kvm/booke.c
> > @@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  {
> >  	struct debug_reg *dbg_reg;
> >  	int n, b = 0, w = 0;
> > +	int ret = 0;
> > +
> > +	vcpu_load(vcpu);
> >  
> >  	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
> >  		vcpu->arch.dbg_reg.dbcr0 = 0;
> >  		vcpu->guest_debug = 0;
> >  		kvm_guest_protect_msr(vcpu, MSR_DE, false);
> > -		return 0;
> > +		goto out;
> >  	}
> >  
> >  	kvm_guest_protect_msr(vcpu, MSR_DE, true);
> > @@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  #endif
> >  
> >  	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
> > -		return 0;
> > +		goto out;
> >  
> > +	ret = -EINVAL;
> >  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
> >  		uint64_t addr = dbg->arch.bp[n].addr;
> >  		uint32_t type = dbg->arch.bp[n].type;
> > @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
> >  			     KVMPPC_DEBUG_WATCH_WRITE |
> >  			     KVMPPC_DEBUG_BREAKPOINT))
> > -			return -EINVAL;
> > +			goto out;
> >  
> >  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
> >  			/* Setting H/W breakpoint */
> >  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
> > -				return -EINVAL;
> > +				goto out;
> >  		} else {
> >  			/* Setting H/W watchpoint */
> >  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
> >  							type, w++))
> > -				return -EINVAL;
> > +				goto out;
> >  		}
> >  	}
> >  
> > -	return 0;
> > +	ret = 0;
> 
> I would probably set the -EINVAL in the individual branches (so it is
> clear that something is wrong, and it is not just a benign exit as in
> the cases above), but your code is correct as well.

I think that's better as well actually.  I got probably got a little
used to that pattern after looking the main dispatcher function for a
while.  I'm happy to change it.

> > +out:
> > +	vcpu_put(vcpu);
> > +	return ret;
> >  }
> >  
> >  void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
> 
> In any case,
> 
> Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Thanks!
-Christoffer

Re: [PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Christoffer Dall]

On Mon, Dec 11, 2017 at 01:39:43PM +0100, Cornelia Huck wrote:
> On Mon,  4 Dec 2017 21:35:33 +0100
> Christoffer Dall <cdall@kernel.org> wrote:
> 
> > From: Christoffer Dall <christoffer.dall@linaro.org>
> > 
> > Move vcpu_load() and vcpu_put() into the architecture specific
> > implementations of kvm_arch_vcpu_ioctl_set_guest_debug().
> > 
> > Reviewed-by: David Hildenbrand <david@redhat.com>
> > Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> > ---
> >  arch/arm64/kvm/guest.c    | 15 ++++++++++++---
> >  arch/powerpc/kvm/book3s.c |  2 ++
> >  arch/powerpc/kvm/booke.c  | 19 +++++++++++++------
> >  arch/s390/kvm/kvm-s390.c  | 16 ++++++++++++----
> >  arch/x86/kvm/x86.c        |  4 +++-
> >  virt/kvm/kvm_main.c       |  2 --
> >  6 files changed, 42 insertions(+), 16 deletions(-)
> > 
> 
> > diff --git a/arch/powerpc/kvm/booke.c b/arch/powerpc/kvm/booke.c
> > index 1b491b89cd43..7cb0e2677e60 100644
> > --- a/arch/powerpc/kvm/booke.c
> > +++ b/arch/powerpc/kvm/booke.c
> > @@ -2018,12 +2018,15 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  {
> >  	struct debug_reg *dbg_reg;
> >  	int n, b = 0, w = 0;
> > +	int ret = 0;
> > +
> > +	vcpu_load(vcpu);
> >  
> >  	if (!(dbg->control & KVM_GUESTDBG_ENABLE)) {
> >  		vcpu->arch.dbg_reg.dbcr0 = 0;
> >  		vcpu->guest_debug = 0;
> >  		kvm_guest_protect_msr(vcpu, MSR_DE, false);
> > -		return 0;
> > +		goto out;
> >  	}
> >  
> >  	kvm_guest_protect_msr(vcpu, MSR_DE, true);
> > @@ -2055,8 +2058,9 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  #endif
> >  
> >  	if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP))
> > -		return 0;
> > +		goto out;
> >  
> > +	ret = -EINVAL;
> >  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
> >  		uint64_t addr = dbg->arch.bp[n].addr;
> >  		uint32_t type = dbg->arch.bp[n].type;
> > @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
> >  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
> >  			     KVMPPC_DEBUG_WATCH_WRITE |
> >  			     KVMPPC_DEBUG_BREAKPOINT))
> > -			return -EINVAL;
> > +			goto out;
> >  
> >  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
> >  			/* Setting H/W breakpoint */
> >  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
> > -				return -EINVAL;
> > +				goto out;
> >  		} else {
> >  			/* Setting H/W watchpoint */
> >  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
> >  							type, w++))
> > -				return -EINVAL;
> > +				goto out;
> >  		}
> >  	}
> >  
> > -	return 0;
> > +	ret = 0;
> 
> I would probably set the -EINVAL in the individual branches (so it is
> clear that something is wrong, and it is not just a benign exit as in
> the cases above), but your code is correct as well.

I think that's better as well actually.  I got probably got a little
used to that pattern after looking the main dispatcher function for a
while.  I'm happy to change it.

> > +out:
> > +	vcpu_put(vcpu);
> > +	return ret;
> >  }
> >  
> >  void kvmppc_booke_vcpu_load(struct kvm_vcpu *vcpu, int cpu)
> 
> In any case,
> 
> Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Thanks!
-Christoffer

Re: [PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Paolo Bonzini]

On 11/12/2017 13:39, Cornelia Huck wrote:
>> +	ret = -EINVAL;
>>  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
>>  		uint64_t addr = dbg->arch.bp[n].addr;
>>  		uint32_t type = dbg->arch.bp[n].type;
>> @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>>  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
>>  			     KVMPPC_DEBUG_WATCH_WRITE |
>>  			     KVMPPC_DEBUG_BREAKPOINT))
>> -			return -EINVAL;
>> +			goto out;
>>  
>>  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
>>  			/* Setting H/W breakpoint */
>>  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
>> -				return -EINVAL;
>> +				goto out;
>>  		} else {
>>  			/* Setting H/W watchpoint */
>>  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
>>  							type, w++))
>> -				return -EINVAL;
>> +				goto out;
>>  		}
>>  	}
>>  
>> -	return 0;
>> +	ret = 0;
> 
> I would probably set the -EINVAL in the individual branches (so it is
> clear that something is wrong, and it is not just a benign exit as in
> the cases above), but your code is correct as well. Let the powerpc
> folks decide.

The idiom that Christoffer used is found elsewhere in KVM, so I'm
accepting his version.  Thanks for the review!

Paolo

Re: [PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Paolo Bonzini]

On 11/12/2017 13:39, Cornelia Huck wrote:
>> +	ret = -EINVAL;
>>  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
>>  		uint64_t addr = dbg->arch.bp[n].addr;
>>  		uint32_t type = dbg->arch.bp[n].type;
>> @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>>  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
>>  			     KVMPPC_DEBUG_WATCH_WRITE |
>>  			     KVMPPC_DEBUG_BREAKPOINT))
>> -			return -EINVAL;
>> +			goto out;
>>  
>>  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
>>  			/* Setting H/W breakpoint */
>>  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
>> -				return -EINVAL;
>> +				goto out;
>>  		} else {
>>  			/* Setting H/W watchpoint */
>>  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
>>  							type, w++))
>> -				return -EINVAL;
>> +				goto out;
>>  		}
>>  	}
>>  
>> -	return 0;
>> +	ret = 0;
> 
> I would probably set the -EINVAL in the individual branches (so it is
> clear that something is wrong, and it is not just a benign exit as in
> the cases above), but your code is correct as well. Let the powerpc
> folks decide.

The idiom that Christoffer used is found elsewhere in KVM, so I'm
accepting his version.  Thanks for the review!

Paolo

[PATCH v3 11/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_guest_debug

[Paolo Bonzini]

On 11/12/2017 13:39, Cornelia Huck wrote:
>> +	ret = -EINVAL;
>>  	for (n = 0; n < (KVMPPC_BOOKE_IAC_NUM + KVMPPC_BOOKE_DAC_NUM); n++) {
>>  		uint64_t addr = dbg->arch.bp[n].addr;
>>  		uint32_t type = dbg->arch.bp[n].type;
>> @@ -2067,21 +2071,24 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
>>  		if (type & ~(KVMPPC_DEBUG_WATCH_READ |
>>  			     KVMPPC_DEBUG_WATCH_WRITE |
>>  			     KVMPPC_DEBUG_BREAKPOINT))
>> -			return -EINVAL;
>> +			goto out;
>>  
>>  		if (type & KVMPPC_DEBUG_BREAKPOINT) {
>>  			/* Setting H/W breakpoint */
>>  			if (kvmppc_booke_add_breakpoint(dbg_reg, addr, b++))
>> -				return -EINVAL;
>> +				goto out;
>>  		} else {
>>  			/* Setting H/W watchpoint */
>>  			if (kvmppc_booke_add_watchpoint(dbg_reg, addr,
>>  							type, w++))
>> -				return -EINVAL;
>> +				goto out;
>>  		}
>>  	}
>>  
>> -	return 0;
>> +	ret = 0;
> 
> I would probably set the -EINVAL in the individual branches (so it is
> clear that something is wrong, and it is not just a benign exit as in
> the cases above), but your code is correct as well. Let the powerpc
> folks decide.

The idiom that Christoffer used is found elsewhere in KVM, so I'm
accepting his version.  Thanks for the review!

Paolo

[PATCH v3 12/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 4 ++++
 arch/x86/kvm/x86.c       | 7 +++++--
 virt/kvm/kvm_main.c      | 2 --
 3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 4bf80b57b5c1..88dcb89656be 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2765,6 +2765,8 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
+	vcpu_load(vcpu);
+
 	/* make sure we have the latest values */
 	save_fpu_regs();
 	if (MACHINE_HAS_VX)
@@ -2773,6 +2775,8 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	else
 		memcpy(fpu->fprs, vcpu->run->s.regs.fprs, sizeof(fpu->fprs));
 	fpu->fpc = vcpu->run->s.regs.fpc;
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 5d19caee6d51..19b70e016858 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7678,9 +7678,11 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	struct fxregs_state *fxsave =
-			&vcpu->arch.guest_fpu.state.fxsave;
+	struct fxregs_state *fxsave;
 
+	vcpu_load(vcpu);
+
+	fxsave = &vcpu->arch.guest_fpu.state.fxsave;
 	memcpy(fpu->fpr, fxsave->st_space, 128);
 	fpu->fcw = fxsave->cwd;
 	fpu->fsw = fxsave->swd;
@@ -7690,6 +7692,7 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	fpu->last_dp = fxsave->rdp;
 	memcpy(fpu->xmm, fxsave->xmm_space, sizeof fxsave->xmm_space);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index c688eb777bec..73ad70af6b2d 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2673,9 +2673,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!fpu)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_fpu(vcpu, fpu);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 12/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 4 ++++
 arch/x86/kvm/x86.c       | 7 +++++--
 virt/kvm/kvm_main.c      | 2 --
 3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 4bf80b57b5c1..88dcb89656be 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2765,6 +2765,8 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
+	vcpu_load(vcpu);
+
 	/* make sure we have the latest values */
 	save_fpu_regs();
 	if (MACHINE_HAS_VX)
@@ -2773,6 +2775,8 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	else
 		memcpy(fpu->fprs, vcpu->run->s.regs.fprs, sizeof(fpu->fprs));
 	fpu->fpc = vcpu->run->s.regs.fpc;
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 5d19caee6d51..19b70e016858 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7678,9 +7678,11 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	struct fxregs_state *fxsave -			&vcpu->arch.guest_fpu.state.fxsave;
+	struct fxregs_state *fxsave;
 
+	vcpu_load(vcpu);
+
+	fxsave = &vcpu->arch.guest_fpu.state.fxsave;
 	memcpy(fpu->fpr, fxsave->st_space, 128);
 	fpu->fcw = fxsave->cwd;
 	fpu->fsw = fxsave->swd;
@@ -7690,6 +7692,7 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	fpu->last_dp = fxsave->rdp;
 	memcpy(fpu->xmm, fxsave->xmm_space, sizeof fxsave->xmm_space);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index c688eb777bec..73ad70af6b2d 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2673,9 +2673,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!fpu)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_fpu(vcpu, fpu);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 12/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 4 ++++
 arch/x86/kvm/x86.c       | 7 +++++--
 virt/kvm/kvm_main.c      | 2 --
 3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 4bf80b57b5c1..88dcb89656be 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2765,6 +2765,8 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
+	vcpu_load(vcpu);
+
 	/* make sure we have the latest values */
 	save_fpu_regs();
 	if (MACHINE_HAS_VX)
@@ -2773,6 +2775,8 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	else
 		memcpy(fpu->fprs, vcpu->run->s.regs.fprs, sizeof(fpu->fprs));
 	fpu->fpc = vcpu->run->s.regs.fpc;
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 5d19caee6d51..19b70e016858 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7678,9 +7678,11 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	struct fxregs_state *fxsave =
-			&vcpu->arch.guest_fpu.state.fxsave;
+	struct fxregs_state *fxsave;
 
+	vcpu_load(vcpu);
+
+	fxsave = &vcpu->arch.guest_fpu.state.fxsave;
 	memcpy(fpu->fpr, fxsave->st_space, 128);
 	fpu->fcw = fxsave->cwd;
 	fpu->fsw = fxsave->swd;
@@ -7690,6 +7692,7 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	fpu->last_dp = fxsave->rdp;
 	memcpy(fpu->xmm, fxsave->xmm_space, sizeof fxsave->xmm_space);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index c688eb777bec..73ad70af6b2d 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2673,9 +2673,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!fpu)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_fpu(vcpu, fpu);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

[PATCH v3 12/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_get_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 4 ++++
 arch/x86/kvm/x86.c       | 7 +++++--
 virt/kvm/kvm_main.c      | 2 --
 3 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 4bf80b57b5c1..88dcb89656be 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2765,6 +2765,8 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
+	vcpu_load(vcpu);
+
 	/* make sure we have the latest values */
 	save_fpu_regs();
 	if (MACHINE_HAS_VX)
@@ -2773,6 +2775,8 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	else
 		memcpy(fpu->fprs, vcpu->run->s.regs.fprs, sizeof(fpu->fprs));
 	fpu->fpc = vcpu->run->s.regs.fpc;
+
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 5d19caee6d51..19b70e016858 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7678,9 +7678,11 @@ int kvm_arch_vcpu_ioctl_translate(struct kvm_vcpu *vcpu,
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	struct fxregs_state *fxsave =
-			&vcpu->arch.guest_fpu.state.fxsave;
+	struct fxregs_state *fxsave;
 
+	vcpu_load(vcpu);
+
+	fxsave = &vcpu->arch.guest_fpu.state.fxsave;
 	memcpy(fpu->fpr, fxsave->st_space, 128);
 	fpu->fcw = fxsave->cwd;
 	fpu->fsw = fxsave->swd;
@@ -7690,6 +7692,7 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	fpu->last_dp = fxsave->rdp;
 	memcpy(fpu->xmm, fxsave->xmm_space, sizeof fxsave->xmm_space);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index c688eb777bec..73ad70af6b2d 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2673,9 +2673,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		r = -ENOMEM;
 		if (!fpu)
 			goto out;
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_get_fpu(vcpu, fpu);
-		vcpu_put(vcpu);
 		if (r)
 			goto out;
 		r = -EFAULT;
-- 
2.14.2

Re: [PATCH v3 12/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:34 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_fpu().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 4 ++++
>  arch/x86/kvm/x86.c       | 7 +++++--
>  virt/kvm/kvm_main.c      | 2 --
>  3 files changed, 9 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 12/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:34 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_fpu().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 4 ++++
>  arch/x86/kvm/x86.c       | 7 +++++--
>  virt/kvm/kvm_main.c      | 2 --
>  3 files changed, 9 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 12/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:34 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_fpu().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 4 ++++
>  arch/x86/kvm/x86.c       | 7 +++++--
>  virt/kvm/kvm_main.c      | 2 --
>  3 files changed, 9 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 12/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_get_fpu

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:34 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_get_fpu().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 4 ++++
>  arch/x86/kvm/x86.c       | 7 +++++--
>  virt/kvm/kvm_main.c      | 2 --
>  3 files changed, 9 insertions(+), 4 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 13/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 15 ++++++++++++---
 arch/x86/kvm/x86.c       |  8 ++++++--
 virt/kvm/kvm_main.c      |  2 --
 3 files changed, 18 insertions(+), 7 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 88dcb89656be..43278f334ce3 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2752,15 +2752,24 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 
 int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	if (test_fp_ctl(fpu->fpc))
-		return -EINVAL;
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
+	if (test_fp_ctl(fpu->fpc)) {
+		ret = -EINVAL;
+		goto out;
+	}
 	vcpu->run->s.regs.fpc = fpu->fpc;
 	if (MACHINE_HAS_VX)
 		convert_fp_to_vx((__vector128 *) vcpu->run->s.regs.vrs,
 				 (freg_t *) fpu->fprs);
 	else
 		memcpy(vcpu->run->s.regs.fprs, &fpu->fprs, sizeof(fpu->fprs));
-	return 0;
+
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 19b70e016858..95a329580c8b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7698,8 +7698,11 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 
 int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	struct fxregs_state *fxsave =
-			&vcpu->arch.guest_fpu.state.fxsave;
+	struct fxregs_state *fxsave;
+
+	vcpu_load(vcpu);
+
+	fxsave = &vcpu->arch.guest_fpu.state.fxsave;
 
 	memcpy(fxsave->st_space, fpu->fpr, 128);
 	fxsave->cwd = fpu->fcw;
@@ -7710,6 +7713,7 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	fxsave->rdp = fpu->last_dp;
 	memcpy(fxsave->xmm_space, fpu->xmm, sizeof fxsave->xmm_space);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 73ad70af6b2d..06751bbecd58 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2689,9 +2689,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			fpu = NULL;
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
-		vcpu_put(vcpu);
 		break;
 	}
 	default:
-- 
2.14.2

[PATCH v3 13/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 15 ++++++++++++---
 arch/x86/kvm/x86.c       |  8 ++++++--
 virt/kvm/kvm_main.c      |  2 --
 3 files changed, 18 insertions(+), 7 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 88dcb89656be..43278f334ce3 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2752,15 +2752,24 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 
 int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	if (test_fp_ctl(fpu->fpc))
-		return -EINVAL;
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
+	if (test_fp_ctl(fpu->fpc)) {
+		ret = -EINVAL;
+		goto out;
+	}
 	vcpu->run->s.regs.fpc = fpu->fpc;
 	if (MACHINE_HAS_VX)
 		convert_fp_to_vx((__vector128 *) vcpu->run->s.regs.vrs,
 				 (freg_t *) fpu->fprs);
 	else
 		memcpy(vcpu->run->s.regs.fprs, &fpu->fprs, sizeof(fpu->fprs));
-	return 0;
+
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 19b70e016858..95a329580c8b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7698,8 +7698,11 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 
 int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	struct fxregs_state *fxsave -			&vcpu->arch.guest_fpu.state.fxsave;
+	struct fxregs_state *fxsave;
+
+	vcpu_load(vcpu);
+
+	fxsave = &vcpu->arch.guest_fpu.state.fxsave;
 
 	memcpy(fxsave->st_space, fpu->fpr, 128);
 	fxsave->cwd = fpu->fcw;
@@ -7710,6 +7713,7 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	fxsave->rdp = fpu->last_dp;
 	memcpy(fxsave->xmm_space, fpu->xmm, sizeof fxsave->xmm_space);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 73ad70af6b2d..06751bbecd58 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2689,9 +2689,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			fpu = NULL;
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
-		vcpu_put(vcpu);
 		break;
 	}
 	default:
-- 
2.14.2

[PATCH v3 13/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 15 ++++++++++++---
 arch/x86/kvm/x86.c       |  8 ++++++--
 virt/kvm/kvm_main.c      |  2 --
 3 files changed, 18 insertions(+), 7 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 88dcb89656be..43278f334ce3 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2752,15 +2752,24 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 
 int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	if (test_fp_ctl(fpu->fpc))
-		return -EINVAL;
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
+	if (test_fp_ctl(fpu->fpc)) {
+		ret = -EINVAL;
+		goto out;
+	}
 	vcpu->run->s.regs.fpc = fpu->fpc;
 	if (MACHINE_HAS_VX)
 		convert_fp_to_vx((__vector128 *) vcpu->run->s.regs.vrs,
 				 (freg_t *) fpu->fprs);
 	else
 		memcpy(vcpu->run->s.regs.fprs, &fpu->fprs, sizeof(fpu->fprs));
-	return 0;
+
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 19b70e016858..95a329580c8b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7698,8 +7698,11 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 
 int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	struct fxregs_state *fxsave =
-			&vcpu->arch.guest_fpu.state.fxsave;
+	struct fxregs_state *fxsave;
+
+	vcpu_load(vcpu);
+
+	fxsave = &vcpu->arch.guest_fpu.state.fxsave;
 
 	memcpy(fxsave->st_space, fpu->fpr, 128);
 	fxsave->cwd = fpu->fcw;
@@ -7710,6 +7713,7 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	fxsave->rdp = fpu->last_dp;
 	memcpy(fxsave->xmm_space, fpu->xmm, sizeof fxsave->xmm_space);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 73ad70af6b2d..06751bbecd58 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2689,9 +2689,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			fpu = NULL;
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
-		vcpu_put(vcpu);
 		break;
 	}
 	default:
-- 
2.14.2

[PATCH v3 13/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move vcpu_load() and vcpu_put() into the architecture specific
implementations of kvm_arch_vcpu_ioctl_set_fpu().

Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/s390/kvm/kvm-s390.c | 15 ++++++++++++---
 arch/x86/kvm/x86.c       |  8 ++++++--
 virt/kvm/kvm_main.c      |  2 --
 3 files changed, 18 insertions(+), 7 deletions(-)

diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 88dcb89656be..43278f334ce3 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -2752,15 +2752,24 @@ int kvm_arch_vcpu_ioctl_get_sregs(struct kvm_vcpu *vcpu,
 
 int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	if (test_fp_ctl(fpu->fpc))
-		return -EINVAL;
+	int ret = 0;
+
+	vcpu_load(vcpu);
+
+	if (test_fp_ctl(fpu->fpc)) {
+		ret = -EINVAL;
+		goto out;
+	}
 	vcpu->run->s.regs.fpc = fpu->fpc;
 	if (MACHINE_HAS_VX)
 		convert_fp_to_vx((__vector128 *) vcpu->run->s.regs.vrs,
 				 (freg_t *) fpu->fprs);
 	else
 		memcpy(vcpu->run->s.regs.fprs, &fpu->fprs, sizeof(fpu->fprs));
-	return 0;
+
+out:
+	vcpu_put(vcpu);
+	return ret;
 }
 
 int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 19b70e016858..95a329580c8b 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -7698,8 +7698,11 @@ int kvm_arch_vcpu_ioctl_get_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 
 int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 {
-	struct fxregs_state *fxsave =
-			&vcpu->arch.guest_fpu.state.fxsave;
+	struct fxregs_state *fxsave;
+
+	vcpu_load(vcpu);
+
+	fxsave = &vcpu->arch.guest_fpu.state.fxsave;
 
 	memcpy(fxsave->st_space, fpu->fpr, 128);
 	fxsave->cwd = fpu->fcw;
@@ -7710,6 +7713,7 @@ int kvm_arch_vcpu_ioctl_set_fpu(struct kvm_vcpu *vcpu, struct kvm_fpu *fpu)
 	fxsave->rdp = fpu->last_dp;
 	memcpy(fxsave->xmm_space, fpu->xmm, sizeof fxsave->xmm_space);
 
+	vcpu_put(vcpu);
 	return 0;
 }
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 73ad70af6b2d..06751bbecd58 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2689,9 +2689,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 			fpu = NULL;
 			goto out;
 		}
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl_set_fpu(vcpu, fpu);
-		vcpu_put(vcpu);
 		break;
 	}
 	default:
-- 
2.14.2

Re: [PATCH v3 13/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:35 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_fpu().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 15 ++++++++++++---
>  arch/x86/kvm/x86.c       |  8 ++++++--
>  virt/kvm/kvm_main.c      |  2 --
>  3 files changed, 18 insertions(+), 7 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

Re: [PATCH v3 13/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:35 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_fpu().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 15 ++++++++++++---
>  arch/x86/kvm/x86.c       |  8 ++++++--
>  virt/kvm/kvm_main.c      |  2 --
>  3 files changed, 18 insertions(+), 7 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 13/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl_set_fpu

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:35 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move vcpu_load() and vcpu_put() into the architecture specific
> implementations of kvm_arch_vcpu_ioctl_set_fpu().
> 
> Reviewed-by: David Hildenbrand <david@redhat.com>
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/s390/kvm/kvm-s390.c | 15 ++++++++++++---
>  arch/x86/kvm/x86.c       |  8 ++++++--
>  virt/kvm/kvm_main.c      |  2 --
>  3 files changed, 18 insertions(+), 7 deletions(-)

Reviewed-by: Cornelia Huck <cohuck@redhat.com>

[PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move the calls to vcpu_load() and vcpu_put() in to the architecture
specific implementations of kvm_arch_vcpu_ioctl() which dispatches
further architecture-specific ioctls on to other functions.

Some architectures support asynchronous vcpu ioctls which cannot call
vcpu_load() or take the vcpu->mutex, because that would prevent
concurrent execution with a running VCPU, which is the intended purpose
of these ioctls, for example because they inject interrupts.

We repeat the separate checks for these specifics in the architecture
code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
calling vcpu_load for these ioctls.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
 arch/powerpc/kvm/powerpc.c | 13 ++++++-----
 arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
 arch/x86/kvm/x86.c         | 22 +++++++++++++-----
 virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
 virt/kvm/kvm_main.c        |  2 --
 6 files changed, 103 insertions(+), 60 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index 3a898712d6cd..4a039341dc29 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
 	void __user *argp = (void __user *)arg;
 	long r;
 
+	if (ioctl == KVM_INTERRUPT) {
+		struct kvm_mips_interrupt irq;
+
+		if (copy_from_user(&irq, argp, sizeof(irq)))
+			return -EFAULT;
+		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
+			  irq.irq);
+
+		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
+	}
+
+	vcpu_load(vcpu);
+
 	switch (ioctl) {
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg, argp, sizeof(reg)))
-			return -EFAULT;
+			break;
 		if (ioctl == KVM_SET_ONE_REG)
-			return kvm_mips_set_reg(vcpu, &reg);
+			r = kvm_mips_set_reg(vcpu, &reg);
 		else
-			return kvm_mips_get_reg(vcpu, &reg);
+			r = kvm_mips_get_reg(vcpu, &reg);
+		break;
 	}
 	case KVM_GET_REG_LIST: {
 		struct kvm_reg_list __user *user_list = argp;
 		struct kvm_reg_list reg_list;
 		unsigned n;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg_list, user_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
 		n = reg_list.n;
 		reg_list.n = kvm_mips_num_regs(vcpu);
 		if (copy_to_user(user_list, &reg_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
+		r = -E2BIG;
 		if (n < reg_list.n)
-			return -E2BIG;
-		return kvm_mips_copy_reg_indices(vcpu, user_list->reg);
-	}
-	case KVM_INTERRUPT:
-		{
-			struct kvm_mips_interrupt irq;
-
-			if (copy_from_user(&irq, argp, sizeof(irq)))
-				return -EFAULT;
-			kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
-				  irq.irq);
-
-			r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
 			break;
-		}
+		r = kvm_mips_copy_reg_indices(vcpu, user_list->reg);
+		break;
+	}
 	case KVM_ENABLE_CAP: {
 		struct kvm_enable_cap cap;
 
+		r = -EFAULT;
 		if (copy_from_user(&cap, argp, sizeof(cap)))
-			return -EFAULT;
+			break;
 		r = kvm_vcpu_ioctl_enable_cap(vcpu, &cap);
 		break;
 	}
 	default:
 		r = -ENOIOCTLCMD;
 	}
+
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index c06bc9552438..6b5dd3a25fe8 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	void __user *argp = (void __user *)arg;
 	long r;
 
-	switch (ioctl) {
-	case KVM_INTERRUPT: {
+	if (ioctl == KVM_INTERRUPT) {
 		struct kvm_interrupt irq;
-		r = -EFAULT;
 		if (copy_from_user(&irq, argp, sizeof(irq)))
-			goto out;
-		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
-		goto out;
+			return -EFAULT;
+		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
 	}
 
+	vcpu_load(vcpu);
+
+	switch (ioctl) {
 	case KVM_ENABLE_CAP:
 	{
 		struct kvm_enable_cap cap;
@@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 
 out:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 43278f334ce3..cd067b63d77f 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	case KVM_S390_IRQ: {
 		struct kvm_s390_irq s390irq;
 
-		r = -EFAULT;
 		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
-			break;
-		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
-		break;
+			return -EFAULT;
+		return kvm_s390_inject_vcpu(vcpu, &s390irq);
 	}
 	case KVM_S390_INTERRUPT: {
 		struct kvm_s390_interrupt s390int;
 		struct kvm_s390_irq s390irq;
 
-		r = -EFAULT;
 		if (copy_from_user(&s390int, argp, sizeof(s390int)))
-			break;
+			return -EFAULT;
 		if (s390int_to_s390irq(&s390int, &s390irq))
 			return -EINVAL;
-		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
-		break;
+		return kvm_s390_inject_vcpu(vcpu, &s390irq);
 	}
+	}
+
+	vcpu_load(vcpu);
+
+	switch (ioctl) {
 	case KVM_S390_STORE_STATUS:
 		idx = srcu_read_lock(&vcpu->kvm->srcu);
 		r = kvm_s390_vcpu_store_status(vcpu, arg);
@@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	default:
 		r = -ENOTTY;
 	}
+
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 95a329580c8b..e35d9d340d7f 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3458,6 +3458,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		void *buffer;
 	} u;
 
+	vcpu_load(vcpu);
+
 	u.buffer = NULL;
 	switch (ioctl) {
 	case KVM_GET_LAPIC: {
@@ -3483,8 +3485,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		if (!lapic_in_kernel(vcpu))
 			goto out;
 		u.lapic = memdup_user(argp, sizeof(*u.lapic));
-		if (IS_ERR(u.lapic))
-			return PTR_ERR(u.lapic);
+		if (IS_ERR(u.lapic)) {
+			r = PTR_ERR(u.lapic);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_set_lapic(vcpu, u.lapic);
 		break;
@@ -3658,8 +3662,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XSAVE: {
 		u.xsave = memdup_user(argp, sizeof(*u.xsave));
-		if (IS_ERR(u.xsave))
-			return PTR_ERR(u.xsave);
+		if (IS_ERR(u.xsave)) {
+			r = PTR_ERR(u.xsave);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_x86_set_xsave(vcpu, u.xsave);
 		break;
@@ -3681,8 +3687,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XCRS: {
 		u.xcrs = memdup_user(argp, sizeof(*u.xcrs));
-		if (IS_ERR(u.xcrs))
-			return PTR_ERR(u.xcrs);
+		if (IS_ERR(u.xcrs)) {
+			r = PTR_ERR(u.xcrs);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_x86_set_xcrs(vcpu, u.xcrs);
 		break;
@@ -3726,6 +3734,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 out:
 	kfree(u.buffer);
+out_nofree:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 9a3acbcf542c..8223c59be507 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -1001,66 +1001,88 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	struct kvm_vcpu *vcpu = filp->private_data;
 	void __user *argp = (void __user *)arg;
 	struct kvm_device_attr attr;
+	long r;
+
+	vcpu_load(vcpu);
 
 	switch (ioctl) {
 	case KVM_ARM_VCPU_INIT: {
 		struct kvm_vcpu_init init;
 
+		r = -EFAULT;
 		if (copy_from_user(&init, argp, sizeof(init)))
-			return -EFAULT;
+			break;
 
-		return kvm_arch_vcpu_ioctl_vcpu_init(vcpu, &init);
+		r = kvm_arch_vcpu_ioctl_vcpu_init(vcpu, &init);
+		break;
 	}
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
 
+		r = -ENOEXEC;
 		if (unlikely(!kvm_vcpu_initialized(vcpu)))
-			return -ENOEXEC;
+			break;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg, argp, sizeof(reg)))
-			return -EFAULT;
+			break;
+
 		if (ioctl == KVM_SET_ONE_REG)
-			return kvm_arm_set_reg(vcpu, &reg);
+			r = kvm_arm_set_reg(vcpu, &reg);
 		else
-			return kvm_arm_get_reg(vcpu, &reg);
+			r = kvm_arm_get_reg(vcpu, &reg);
+		break;
 	}
 	case KVM_GET_REG_LIST: {
 		struct kvm_reg_list __user *user_list = argp;
 		struct kvm_reg_list reg_list;
 		unsigned n;
 
+		r = -ENOEXEC;
 		if (unlikely(!kvm_vcpu_initialized(vcpu)))
-			return -ENOEXEC;
+			break;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg_list, user_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
 		n = reg_list.n;
 		reg_list.n = kvm_arm_num_regs(vcpu);
 		if (copy_to_user(user_list, &reg_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
+		r = -E2BIG;
 		if (n < reg_list.n)
-			return -E2BIG;
-		return kvm_arm_copy_reg_indices(vcpu, user_list->reg);
+			break;
+		r = kvm_arm_copy_reg_indices(vcpu, user_list->reg);
+		break;
 	}
 	case KVM_SET_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_set_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_set_attr(vcpu, &attr);
+		break;
 	}
 	case KVM_GET_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_get_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_get_attr(vcpu, &attr);
+		break;
 	}
 	case KVM_HAS_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_has_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_has_attr(vcpu, &attr);
+		break;
 	}
 	default:
-		return -EINVAL;
+		r = -EINVAL;
 	}
+
+	vcpu_put(vcpu);
+	return r;
 }
 
 /**
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 06751bbecd58..ad5f83159a15 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		break;
 	}
 	default:
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
-		vcpu_put(vcpu);
 	}
 out:
 	mutex_unlock(&vcpu->mutex);
-- 
2.14.2

[PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move the calls to vcpu_load() and vcpu_put() in to the architecture
specific implementations of kvm_arch_vcpu_ioctl() which dispatches
further architecture-specific ioctls on to other functions.

Some architectures support asynchronous vcpu ioctls which cannot call
vcpu_load() or take the vcpu->mutex, because that would prevent
concurrent execution with a running VCPU, which is the intended purpose
of these ioctls, for example because they inject interrupts.

We repeat the separate checks for these specifics in the architecture
code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
calling vcpu_load for these ioctls.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
 arch/powerpc/kvm/powerpc.c | 13 ++++++-----
 arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
 arch/x86/kvm/x86.c         | 22 +++++++++++++-----
 virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
 virt/kvm/kvm_main.c        |  2 --
 6 files changed, 103 insertions(+), 60 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index 3a898712d6cd..4a039341dc29 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
 	void __user *argp = (void __user *)arg;
 	long r;
 
+	if (ioctl = KVM_INTERRUPT) {
+		struct kvm_mips_interrupt irq;
+
+		if (copy_from_user(&irq, argp, sizeof(irq)))
+			return -EFAULT;
+		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
+			  irq.irq);
+
+		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
+	}
+
+	vcpu_load(vcpu);
+
 	switch (ioctl) {
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg, argp, sizeof(reg)))
-			return -EFAULT;
+			break;
 		if (ioctl = KVM_SET_ONE_REG)
-			return kvm_mips_set_reg(vcpu, &reg);
+			r = kvm_mips_set_reg(vcpu, &reg);
 		else
-			return kvm_mips_get_reg(vcpu, &reg);
+			r = kvm_mips_get_reg(vcpu, &reg);
+		break;
 	}
 	case KVM_GET_REG_LIST: {
 		struct kvm_reg_list __user *user_list = argp;
 		struct kvm_reg_list reg_list;
 		unsigned n;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg_list, user_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
 		n = reg_list.n;
 		reg_list.n = kvm_mips_num_regs(vcpu);
 		if (copy_to_user(user_list, &reg_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
+		r = -E2BIG;
 		if (n < reg_list.n)
-			return -E2BIG;
-		return kvm_mips_copy_reg_indices(vcpu, user_list->reg);
-	}
-	case KVM_INTERRUPT:
-		{
-			struct kvm_mips_interrupt irq;
-
-			if (copy_from_user(&irq, argp, sizeof(irq)))
-				return -EFAULT;
-			kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
-				  irq.irq);
-
-			r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
 			break;
-		}
+		r = kvm_mips_copy_reg_indices(vcpu, user_list->reg);
+		break;
+	}
 	case KVM_ENABLE_CAP: {
 		struct kvm_enable_cap cap;
 
+		r = -EFAULT;
 		if (copy_from_user(&cap, argp, sizeof(cap)))
-			return -EFAULT;
+			break;
 		r = kvm_vcpu_ioctl_enable_cap(vcpu, &cap);
 		break;
 	}
 	default:
 		r = -ENOIOCTLCMD;
 	}
+
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index c06bc9552438..6b5dd3a25fe8 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	void __user *argp = (void __user *)arg;
 	long r;
 
-	switch (ioctl) {
-	case KVM_INTERRUPT: {
+	if (ioctl = KVM_INTERRUPT) {
 		struct kvm_interrupt irq;
-		r = -EFAULT;
 		if (copy_from_user(&irq, argp, sizeof(irq)))
-			goto out;
-		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
-		goto out;
+			return -EFAULT;
+		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
 	}
 
+	vcpu_load(vcpu);
+
+	switch (ioctl) {
 	case KVM_ENABLE_CAP:
 	{
 		struct kvm_enable_cap cap;
@@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 
 out:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 43278f334ce3..cd067b63d77f 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	case KVM_S390_IRQ: {
 		struct kvm_s390_irq s390irq;
 
-		r = -EFAULT;
 		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
-			break;
-		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
-		break;
+			return -EFAULT;
+		return kvm_s390_inject_vcpu(vcpu, &s390irq);
 	}
 	case KVM_S390_INTERRUPT: {
 		struct kvm_s390_interrupt s390int;
 		struct kvm_s390_irq s390irq;
 
-		r = -EFAULT;
 		if (copy_from_user(&s390int, argp, sizeof(s390int)))
-			break;
+			return -EFAULT;
 		if (s390int_to_s390irq(&s390int, &s390irq))
 			return -EINVAL;
-		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
-		break;
+		return kvm_s390_inject_vcpu(vcpu, &s390irq);
 	}
+	}
+
+	vcpu_load(vcpu);
+
+	switch (ioctl) {
 	case KVM_S390_STORE_STATUS:
 		idx = srcu_read_lock(&vcpu->kvm->srcu);
 		r = kvm_s390_vcpu_store_status(vcpu, arg);
@@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	default:
 		r = -ENOTTY;
 	}
+
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 95a329580c8b..e35d9d340d7f 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3458,6 +3458,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		void *buffer;
 	} u;
 
+	vcpu_load(vcpu);
+
 	u.buffer = NULL;
 	switch (ioctl) {
 	case KVM_GET_LAPIC: {
@@ -3483,8 +3485,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		if (!lapic_in_kernel(vcpu))
 			goto out;
 		u.lapic = memdup_user(argp, sizeof(*u.lapic));
-		if (IS_ERR(u.lapic))
-			return PTR_ERR(u.lapic);
+		if (IS_ERR(u.lapic)) {
+			r = PTR_ERR(u.lapic);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_set_lapic(vcpu, u.lapic);
 		break;
@@ -3658,8 +3662,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XSAVE: {
 		u.xsave = memdup_user(argp, sizeof(*u.xsave));
-		if (IS_ERR(u.xsave))
-			return PTR_ERR(u.xsave);
+		if (IS_ERR(u.xsave)) {
+			r = PTR_ERR(u.xsave);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_x86_set_xsave(vcpu, u.xsave);
 		break;
@@ -3681,8 +3687,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XCRS: {
 		u.xcrs = memdup_user(argp, sizeof(*u.xcrs));
-		if (IS_ERR(u.xcrs))
-			return PTR_ERR(u.xcrs);
+		if (IS_ERR(u.xcrs)) {
+			r = PTR_ERR(u.xcrs);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_x86_set_xcrs(vcpu, u.xcrs);
 		break;
@@ -3726,6 +3734,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 out:
 	kfree(u.buffer);
+out_nofree:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 9a3acbcf542c..8223c59be507 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -1001,66 +1001,88 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	struct kvm_vcpu *vcpu = filp->private_data;
 	void __user *argp = (void __user *)arg;
 	struct kvm_device_attr attr;
+	long r;
+
+	vcpu_load(vcpu);
 
 	switch (ioctl) {
 	case KVM_ARM_VCPU_INIT: {
 		struct kvm_vcpu_init init;
 
+		r = -EFAULT;
 		if (copy_from_user(&init, argp, sizeof(init)))
-			return -EFAULT;
+			break;
 
-		return kvm_arch_vcpu_ioctl_vcpu_init(vcpu, &init);
+		r = kvm_arch_vcpu_ioctl_vcpu_init(vcpu, &init);
+		break;
 	}
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
 
+		r = -ENOEXEC;
 		if (unlikely(!kvm_vcpu_initialized(vcpu)))
-			return -ENOEXEC;
+			break;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg, argp, sizeof(reg)))
-			return -EFAULT;
+			break;
+
 		if (ioctl = KVM_SET_ONE_REG)
-			return kvm_arm_set_reg(vcpu, &reg);
+			r = kvm_arm_set_reg(vcpu, &reg);
 		else
-			return kvm_arm_get_reg(vcpu, &reg);
+			r = kvm_arm_get_reg(vcpu, &reg);
+		break;
 	}
 	case KVM_GET_REG_LIST: {
 		struct kvm_reg_list __user *user_list = argp;
 		struct kvm_reg_list reg_list;
 		unsigned n;
 
+		r = -ENOEXEC;
 		if (unlikely(!kvm_vcpu_initialized(vcpu)))
-			return -ENOEXEC;
+			break;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg_list, user_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
 		n = reg_list.n;
 		reg_list.n = kvm_arm_num_regs(vcpu);
 		if (copy_to_user(user_list, &reg_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
+		r = -E2BIG;
 		if (n < reg_list.n)
-			return -E2BIG;
-		return kvm_arm_copy_reg_indices(vcpu, user_list->reg);
+			break;
+		r = kvm_arm_copy_reg_indices(vcpu, user_list->reg);
+		break;
 	}
 	case KVM_SET_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_set_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_set_attr(vcpu, &attr);
+		break;
 	}
 	case KVM_GET_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_get_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_get_attr(vcpu, &attr);
+		break;
 	}
 	case KVM_HAS_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_has_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_has_attr(vcpu, &attr);
+		break;
 	}
 	default:
-		return -EINVAL;
+		r = -EINVAL;
 	}
+
+	vcpu_put(vcpu);
+	return r;
 }
 
 /**
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 06751bbecd58..ad5f83159a15 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		break;
 	}
 	default:
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
-		vcpu_put(vcpu);
 	}
 out:
 	mutex_unlock(&vcpu->mutex);
-- 
2.14.2

[PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move the calls to vcpu_load() and vcpu_put() in to the architecture
specific implementations of kvm_arch_vcpu_ioctl() which dispatches
further architecture-specific ioctls on to other functions.

Some architectures support asynchronous vcpu ioctls which cannot call
vcpu_load() or take the vcpu->mutex, because that would prevent
concurrent execution with a running VCPU, which is the intended purpose
of these ioctls, for example because they inject interrupts.

We repeat the separate checks for these specifics in the architecture
code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
calling vcpu_load for these ioctls.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
 arch/powerpc/kvm/powerpc.c | 13 ++++++-----
 arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
 arch/x86/kvm/x86.c         | 22 +++++++++++++-----
 virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
 virt/kvm/kvm_main.c        |  2 --
 6 files changed, 103 insertions(+), 60 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index 3a898712d6cd..4a039341dc29 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
 	void __user *argp = (void __user *)arg;
 	long r;
 
+	if (ioctl == KVM_INTERRUPT) {
+		struct kvm_mips_interrupt irq;
+
+		if (copy_from_user(&irq, argp, sizeof(irq)))
+			return -EFAULT;
+		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
+			  irq.irq);
+
+		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
+	}
+
+	vcpu_load(vcpu);
+
 	switch (ioctl) {
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg, argp, sizeof(reg)))
-			return -EFAULT;
+			break;
 		if (ioctl == KVM_SET_ONE_REG)
-			return kvm_mips_set_reg(vcpu, &reg);
+			r = kvm_mips_set_reg(vcpu, &reg);
 		else
-			return kvm_mips_get_reg(vcpu, &reg);
+			r = kvm_mips_get_reg(vcpu, &reg);
+		break;
 	}
 	case KVM_GET_REG_LIST: {
 		struct kvm_reg_list __user *user_list = argp;
 		struct kvm_reg_list reg_list;
 		unsigned n;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg_list, user_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
 		n = reg_list.n;
 		reg_list.n = kvm_mips_num_regs(vcpu);
 		if (copy_to_user(user_list, &reg_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
+		r = -E2BIG;
 		if (n < reg_list.n)
-			return -E2BIG;
-		return kvm_mips_copy_reg_indices(vcpu, user_list->reg);
-	}
-	case KVM_INTERRUPT:
-		{
-			struct kvm_mips_interrupt irq;
-
-			if (copy_from_user(&irq, argp, sizeof(irq)))
-				return -EFAULT;
-			kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
-				  irq.irq);
-
-			r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
 			break;
-		}
+		r = kvm_mips_copy_reg_indices(vcpu, user_list->reg);
+		break;
+	}
 	case KVM_ENABLE_CAP: {
 		struct kvm_enable_cap cap;
 
+		r = -EFAULT;
 		if (copy_from_user(&cap, argp, sizeof(cap)))
-			return -EFAULT;
+			break;
 		r = kvm_vcpu_ioctl_enable_cap(vcpu, &cap);
 		break;
 	}
 	default:
 		r = -ENOIOCTLCMD;
 	}
+
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index c06bc9552438..6b5dd3a25fe8 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	void __user *argp = (void __user *)arg;
 	long r;
 
-	switch (ioctl) {
-	case KVM_INTERRUPT: {
+	if (ioctl == KVM_INTERRUPT) {
 		struct kvm_interrupt irq;
-		r = -EFAULT;
 		if (copy_from_user(&irq, argp, sizeof(irq)))
-			goto out;
-		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
-		goto out;
+			return -EFAULT;
+		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
 	}
 
+	vcpu_load(vcpu);
+
+	switch (ioctl) {
 	case KVM_ENABLE_CAP:
 	{
 		struct kvm_enable_cap cap;
@@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 
 out:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 43278f334ce3..cd067b63d77f 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	case KVM_S390_IRQ: {
 		struct kvm_s390_irq s390irq;
 
-		r = -EFAULT;
 		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
-			break;
-		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
-		break;
+			return -EFAULT;
+		return kvm_s390_inject_vcpu(vcpu, &s390irq);
 	}
 	case KVM_S390_INTERRUPT: {
 		struct kvm_s390_interrupt s390int;
 		struct kvm_s390_irq s390irq;
 
-		r = -EFAULT;
 		if (copy_from_user(&s390int, argp, sizeof(s390int)))
-			break;
+			return -EFAULT;
 		if (s390int_to_s390irq(&s390int, &s390irq))
 			return -EINVAL;
-		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
-		break;
+		return kvm_s390_inject_vcpu(vcpu, &s390irq);
 	}
+	}
+
+	vcpu_load(vcpu);
+
+	switch (ioctl) {
 	case KVM_S390_STORE_STATUS:
 		idx = srcu_read_lock(&vcpu->kvm->srcu);
 		r = kvm_s390_vcpu_store_status(vcpu, arg);
@@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	default:
 		r = -ENOTTY;
 	}
+
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 95a329580c8b..e35d9d340d7f 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3458,6 +3458,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		void *buffer;
 	} u;
 
+	vcpu_load(vcpu);
+
 	u.buffer = NULL;
 	switch (ioctl) {
 	case KVM_GET_LAPIC: {
@@ -3483,8 +3485,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		if (!lapic_in_kernel(vcpu))
 			goto out;
 		u.lapic = memdup_user(argp, sizeof(*u.lapic));
-		if (IS_ERR(u.lapic))
-			return PTR_ERR(u.lapic);
+		if (IS_ERR(u.lapic)) {
+			r = PTR_ERR(u.lapic);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_set_lapic(vcpu, u.lapic);
 		break;
@@ -3658,8 +3662,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XSAVE: {
 		u.xsave = memdup_user(argp, sizeof(*u.xsave));
-		if (IS_ERR(u.xsave))
-			return PTR_ERR(u.xsave);
+		if (IS_ERR(u.xsave)) {
+			r = PTR_ERR(u.xsave);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_x86_set_xsave(vcpu, u.xsave);
 		break;
@@ -3681,8 +3687,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XCRS: {
 		u.xcrs = memdup_user(argp, sizeof(*u.xcrs));
-		if (IS_ERR(u.xcrs))
-			return PTR_ERR(u.xcrs);
+		if (IS_ERR(u.xcrs)) {
+			r = PTR_ERR(u.xcrs);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_x86_set_xcrs(vcpu, u.xcrs);
 		break;
@@ -3726,6 +3734,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 out:
 	kfree(u.buffer);
+out_nofree:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 9a3acbcf542c..8223c59be507 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -1001,66 +1001,88 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	struct kvm_vcpu *vcpu = filp->private_data;
 	void __user *argp = (void __user *)arg;
 	struct kvm_device_attr attr;
+	long r;
+
+	vcpu_load(vcpu);
 
 	switch (ioctl) {
 	case KVM_ARM_VCPU_INIT: {
 		struct kvm_vcpu_init init;
 
+		r = -EFAULT;
 		if (copy_from_user(&init, argp, sizeof(init)))
-			return -EFAULT;
+			break;
 
-		return kvm_arch_vcpu_ioctl_vcpu_init(vcpu, &init);
+		r = kvm_arch_vcpu_ioctl_vcpu_init(vcpu, &init);
+		break;
 	}
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
 
+		r = -ENOEXEC;
 		if (unlikely(!kvm_vcpu_initialized(vcpu)))
-			return -ENOEXEC;
+			break;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg, argp, sizeof(reg)))
-			return -EFAULT;
+			break;
+
 		if (ioctl == KVM_SET_ONE_REG)
-			return kvm_arm_set_reg(vcpu, &reg);
+			r = kvm_arm_set_reg(vcpu, &reg);
 		else
-			return kvm_arm_get_reg(vcpu, &reg);
+			r = kvm_arm_get_reg(vcpu, &reg);
+		break;
 	}
 	case KVM_GET_REG_LIST: {
 		struct kvm_reg_list __user *user_list = argp;
 		struct kvm_reg_list reg_list;
 		unsigned n;
 
+		r = -ENOEXEC;
 		if (unlikely(!kvm_vcpu_initialized(vcpu)))
-			return -ENOEXEC;
+			break;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg_list, user_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
 		n = reg_list.n;
 		reg_list.n = kvm_arm_num_regs(vcpu);
 		if (copy_to_user(user_list, &reg_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
+		r = -E2BIG;
 		if (n < reg_list.n)
-			return -E2BIG;
-		return kvm_arm_copy_reg_indices(vcpu, user_list->reg);
+			break;
+		r = kvm_arm_copy_reg_indices(vcpu, user_list->reg);
+		break;
 	}
 	case KVM_SET_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_set_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_set_attr(vcpu, &attr);
+		break;
 	}
 	case KVM_GET_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_get_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_get_attr(vcpu, &attr);
+		break;
 	}
 	case KVM_HAS_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_has_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_has_attr(vcpu, &attr);
+		break;
 	}
 	default:
-		return -EINVAL;
+		r = -EINVAL;
 	}
+
+	vcpu_put(vcpu);
+	return r;
 }
 
 /**
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 06751bbecd58..ad5f83159a15 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		break;
 	}
 	default:
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
-		vcpu_put(vcpu);
 	}
 out:
 	mutex_unlock(&vcpu->mutex);
-- 
2.14.2

[PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Move the calls to vcpu_load() and vcpu_put() in to the architecture
specific implementations of kvm_arch_vcpu_ioctl() which dispatches
further architecture-specific ioctls on to other functions.

Some architectures support asynchronous vcpu ioctls which cannot call
vcpu_load() or take the vcpu->mutex, because that would prevent
concurrent execution with a running VCPU, which is the intended purpose
of these ioctls, for example because they inject interrupts.

We repeat the separate checks for these specifics in the architecture
code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
calling vcpu_load for these ioctls.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
 arch/powerpc/kvm/powerpc.c | 13 ++++++-----
 arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
 arch/x86/kvm/x86.c         | 22 +++++++++++++-----
 virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
 virt/kvm/kvm_main.c        |  2 --
 6 files changed, 103 insertions(+), 60 deletions(-)

diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
index 3a898712d6cd..4a039341dc29 100644
--- a/arch/mips/kvm/mips.c
+++ b/arch/mips/kvm/mips.c
@@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
 	void __user *argp = (void __user *)arg;
 	long r;
 
+	if (ioctl == KVM_INTERRUPT) {
+		struct kvm_mips_interrupt irq;
+
+		if (copy_from_user(&irq, argp, sizeof(irq)))
+			return -EFAULT;
+		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
+			  irq.irq);
+
+		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
+	}
+
+	vcpu_load(vcpu);
+
 	switch (ioctl) {
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg, argp, sizeof(reg)))
-			return -EFAULT;
+			break;
 		if (ioctl == KVM_SET_ONE_REG)
-			return kvm_mips_set_reg(vcpu, &reg);
+			r = kvm_mips_set_reg(vcpu, &reg);
 		else
-			return kvm_mips_get_reg(vcpu, &reg);
+			r = kvm_mips_get_reg(vcpu, &reg);
+		break;
 	}
 	case KVM_GET_REG_LIST: {
 		struct kvm_reg_list __user *user_list = argp;
 		struct kvm_reg_list reg_list;
 		unsigned n;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg_list, user_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
 		n = reg_list.n;
 		reg_list.n = kvm_mips_num_regs(vcpu);
 		if (copy_to_user(user_list, &reg_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
+		r = -E2BIG;
 		if (n < reg_list.n)
-			return -E2BIG;
-		return kvm_mips_copy_reg_indices(vcpu, user_list->reg);
-	}
-	case KVM_INTERRUPT:
-		{
-			struct kvm_mips_interrupt irq;
-
-			if (copy_from_user(&irq, argp, sizeof(irq)))
-				return -EFAULT;
-			kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
-				  irq.irq);
-
-			r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
 			break;
-		}
+		r = kvm_mips_copy_reg_indices(vcpu, user_list->reg);
+		break;
+	}
 	case KVM_ENABLE_CAP: {
 		struct kvm_enable_cap cap;
 
+		r = -EFAULT;
 		if (copy_from_user(&cap, argp, sizeof(cap)))
-			return -EFAULT;
+			break;
 		r = kvm_vcpu_ioctl_enable_cap(vcpu, &cap);
 		break;
 	}
 	default:
 		r = -ENOIOCTLCMD;
 	}
+
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
index c06bc9552438..6b5dd3a25fe8 100644
--- a/arch/powerpc/kvm/powerpc.c
+++ b/arch/powerpc/kvm/powerpc.c
@@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	void __user *argp = (void __user *)arg;
 	long r;
 
-	switch (ioctl) {
-	case KVM_INTERRUPT: {
+	if (ioctl == KVM_INTERRUPT) {
 		struct kvm_interrupt irq;
-		r = -EFAULT;
 		if (copy_from_user(&irq, argp, sizeof(irq)))
-			goto out;
-		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
-		goto out;
+			return -EFAULT;
+		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
 	}
 
+	vcpu_load(vcpu);
+
+	switch (ioctl) {
 	case KVM_ENABLE_CAP:
 	{
 		struct kvm_enable_cap cap;
@@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 
 out:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 43278f334ce3..cd067b63d77f 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	case KVM_S390_IRQ: {
 		struct kvm_s390_irq s390irq;
 
-		r = -EFAULT;
 		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
-			break;
-		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
-		break;
+			return -EFAULT;
+		return kvm_s390_inject_vcpu(vcpu, &s390irq);
 	}
 	case KVM_S390_INTERRUPT: {
 		struct kvm_s390_interrupt s390int;
 		struct kvm_s390_irq s390irq;
 
-		r = -EFAULT;
 		if (copy_from_user(&s390int, argp, sizeof(s390int)))
-			break;
+			return -EFAULT;
 		if (s390int_to_s390irq(&s390int, &s390irq))
 			return -EINVAL;
-		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
-		break;
+		return kvm_s390_inject_vcpu(vcpu, &s390irq);
 	}
+	}
+
+	vcpu_load(vcpu);
+
+	switch (ioctl) {
 	case KVM_S390_STORE_STATUS:
 		idx = srcu_read_lock(&vcpu->kvm->srcu);
 		r = kvm_s390_vcpu_store_status(vcpu, arg);
@@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	default:
 		r = -ENOTTY;
 	}
+
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 95a329580c8b..e35d9d340d7f 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3458,6 +3458,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		void *buffer;
 	} u;
 
+	vcpu_load(vcpu);
+
 	u.buffer = NULL;
 	switch (ioctl) {
 	case KVM_GET_LAPIC: {
@@ -3483,8 +3485,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		if (!lapic_in_kernel(vcpu))
 			goto out;
 		u.lapic = memdup_user(argp, sizeof(*u.lapic));
-		if (IS_ERR(u.lapic))
-			return PTR_ERR(u.lapic);
+		if (IS_ERR(u.lapic)) {
+			r = PTR_ERR(u.lapic);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_set_lapic(vcpu, u.lapic);
 		break;
@@ -3658,8 +3662,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XSAVE: {
 		u.xsave = memdup_user(argp, sizeof(*u.xsave));
-		if (IS_ERR(u.xsave))
-			return PTR_ERR(u.xsave);
+		if (IS_ERR(u.xsave)) {
+			r = PTR_ERR(u.xsave);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_x86_set_xsave(vcpu, u.xsave);
 		break;
@@ -3681,8 +3687,10 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 	case KVM_SET_XCRS: {
 		u.xcrs = memdup_user(argp, sizeof(*u.xcrs));
-		if (IS_ERR(u.xcrs))
-			return PTR_ERR(u.xcrs);
+		if (IS_ERR(u.xcrs)) {
+			r = PTR_ERR(u.xcrs);
+			goto out_nofree;
+		}
 
 		r = kvm_vcpu_ioctl_x86_set_xcrs(vcpu, u.xcrs);
 		break;
@@ -3726,6 +3734,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	}
 out:
 	kfree(u.buffer);
+out_nofree:
+	vcpu_put(vcpu);
 	return r;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 9a3acbcf542c..8223c59be507 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -1001,66 +1001,88 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	struct kvm_vcpu *vcpu = filp->private_data;
 	void __user *argp = (void __user *)arg;
 	struct kvm_device_attr attr;
+	long r;
+
+	vcpu_load(vcpu);
 
 	switch (ioctl) {
 	case KVM_ARM_VCPU_INIT: {
 		struct kvm_vcpu_init init;
 
+		r = -EFAULT;
 		if (copy_from_user(&init, argp, sizeof(init)))
-			return -EFAULT;
+			break;
 
-		return kvm_arch_vcpu_ioctl_vcpu_init(vcpu, &init);
+		r = kvm_arch_vcpu_ioctl_vcpu_init(vcpu, &init);
+		break;
 	}
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
 
+		r = -ENOEXEC;
 		if (unlikely(!kvm_vcpu_initialized(vcpu)))
-			return -ENOEXEC;
+			break;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg, argp, sizeof(reg)))
-			return -EFAULT;
+			break;
+
 		if (ioctl == KVM_SET_ONE_REG)
-			return kvm_arm_set_reg(vcpu, &reg);
+			r = kvm_arm_set_reg(vcpu, &reg);
 		else
-			return kvm_arm_get_reg(vcpu, &reg);
+			r = kvm_arm_get_reg(vcpu, &reg);
+		break;
 	}
 	case KVM_GET_REG_LIST: {
 		struct kvm_reg_list __user *user_list = argp;
 		struct kvm_reg_list reg_list;
 		unsigned n;
 
+		r = -ENOEXEC;
 		if (unlikely(!kvm_vcpu_initialized(vcpu)))
-			return -ENOEXEC;
+			break;
 
+		r = -EFAULT;
 		if (copy_from_user(&reg_list, user_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
 		n = reg_list.n;
 		reg_list.n = kvm_arm_num_regs(vcpu);
 		if (copy_to_user(user_list, &reg_list, sizeof(reg_list)))
-			return -EFAULT;
+			break;
+		r = -E2BIG;
 		if (n < reg_list.n)
-			return -E2BIG;
-		return kvm_arm_copy_reg_indices(vcpu, user_list->reg);
+			break;
+		r = kvm_arm_copy_reg_indices(vcpu, user_list->reg);
+		break;
 	}
 	case KVM_SET_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_set_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_set_attr(vcpu, &attr);
+		break;
 	}
 	case KVM_GET_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_get_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_get_attr(vcpu, &attr);
+		break;
 	}
 	case KVM_HAS_DEVICE_ATTR: {
+		r = -EFAULT;
 		if (copy_from_user(&attr, argp, sizeof(attr)))
-			return -EFAULT;
-		return kvm_arm_vcpu_has_attr(vcpu, &attr);
+			break;
+		r = kvm_arm_vcpu_has_attr(vcpu, &attr);
+		break;
 	}
 	default:
-		return -EINVAL;
+		r = -EINVAL;
 	}
+
+	vcpu_put(vcpu);
+	return r;
 }
 
 /**
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 06751bbecd58..ad5f83159a15 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
 		break;
 	}
 	default:
-		vcpu_load(vcpu);
 		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
-		vcpu_put(vcpu);
 	}
 out:
 	mutex_unlock(&vcpu->mutex);
-- 
2.14.2

Re: [PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:36 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move the calls to vcpu_load() and vcpu_put() in to the architecture
> specific implementations of kvm_arch_vcpu_ioctl() which dispatches
> further architecture-specific ioctls on to other functions.
> 
> Some architectures support asynchronous vcpu ioctls which cannot call
> vcpu_load() or take the vcpu->mutex, because that would prevent
> concurrent execution with a running VCPU, which is the intended purpose
> of these ioctls, for example because they inject interrupts.
> 
> We repeat the separate checks for these specifics in the architecture
> code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
> calling vcpu_load for these ioctls.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
>  arch/powerpc/kvm/powerpc.c | 13 ++++++-----
>  arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
>  arch/x86/kvm/x86.c         | 22 +++++++++++++-----
>  virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
>  virt/kvm/kvm_main.c        |  2 --
>  6 files changed, 103 insertions(+), 60 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index 3a898712d6cd..4a039341dc29 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
>  	void __user *argp = (void __user *)arg;
>  	long r;
>  
> +	if (ioctl == KVM_INTERRUPT) {

I would add a comment here that this ioctl is async to vcpu execution,
so it is understandable why you skip the vcpu_load().

[As an aside, it is nice that this is now more obvious when looking at
the architectures' handlers.]

> +		struct kvm_mips_interrupt irq;
> +
> +		if (copy_from_user(&irq, argp, sizeof(irq)))
> +			return -EFAULT;
> +		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
> +			  irq.irq);
> +
> +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> +	}
> +
> +	vcpu_load(vcpu);
> +
>  	switch (ioctl) {

(...)

> diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
> index c06bc9552438..6b5dd3a25fe8 100644
> --- a/arch/powerpc/kvm/powerpc.c
> +++ b/arch/powerpc/kvm/powerpc.c
> @@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	void __user *argp = (void __user *)arg;
>  	long r;
>  
> -	switch (ioctl) {
> -	case KVM_INTERRUPT: {
> +	if (ioctl == KVM_INTERRUPT) {

Same here.

>  		struct kvm_interrupt irq;
> -		r = -EFAULT;
>  		if (copy_from_user(&irq, argp, sizeof(irq)))
> -			goto out;
> -		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> -		goto out;
> +			return -EFAULT;
> +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
>  	}
>  
> +	vcpu_load(vcpu);
> +
> +	switch (ioctl) {
>  	case KVM_ENABLE_CAP:
>  	{
>  		struct kvm_enable_cap cap;
> @@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	}
>  
>  out:
> +	vcpu_put(vcpu);
>  	return r;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 43278f334ce3..cd067b63d77f 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	case KVM_S390_IRQ: {
>  		struct kvm_s390_irq s390irq;
>  
> -		r = -EFAULT;
>  		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
> -			break;
> -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> -		break;
> +			return -EFAULT;
> +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
>  	}
>  	case KVM_S390_INTERRUPT: {
>  		struct kvm_s390_interrupt s390int;
>  		struct kvm_s390_irq s390irq;
>  
> -		r = -EFAULT;
>  		if (copy_from_user(&s390int, argp, sizeof(s390int)))
> -			break;
> +			return -EFAULT;
>  		if (s390int_to_s390irq(&s390int, &s390irq))
>  			return -EINVAL;
> -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> -		break;
> +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
>  	}
> +	}

I find the special casing with the immediate return a bit ugly. Maybe
introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
its default case and return here if ret != -ENOIOCTLCMD? Christian,
what do you think?

> +
> +	vcpu_load(vcpu);
> +
> +	switch (ioctl) {
>  	case KVM_S390_STORE_STATUS:
>  		idx = srcu_read_lock(&vcpu->kvm->srcu);
>  		r = kvm_s390_vcpu_store_status(vcpu, arg);
> @@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	default:
>  		r = -ENOTTY;
>  	}
> +
> +	vcpu_put(vcpu);
>  	return r;
>  }
>  

(...)

> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 06751bbecd58..ad5f83159a15 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		break;
>  	}
>  	default:
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
> -		vcpu_put(vcpu);
>  	}
>  out:
>  	mutex_unlock(&vcpu->mutex);

It would be nice if we could get rid of the special casing at the
beginning of this function, but as it would involve not taking the
mutex for special cases (and not releasing it for those special cases),
I don't see an elegant way to do that.

Re: [PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:36 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move the calls to vcpu_load() and vcpu_put() in to the architecture
> specific implementations of kvm_arch_vcpu_ioctl() which dispatches
> further architecture-specific ioctls on to other functions.
> 
> Some architectures support asynchronous vcpu ioctls which cannot call
> vcpu_load() or take the vcpu->mutex, because that would prevent
> concurrent execution with a running VCPU, which is the intended purpose
> of these ioctls, for example because they inject interrupts.
> 
> We repeat the separate checks for these specifics in the architecture
> code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
> calling vcpu_load for these ioctls.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
>  arch/powerpc/kvm/powerpc.c | 13 ++++++-----
>  arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
>  arch/x86/kvm/x86.c         | 22 +++++++++++++-----
>  virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
>  virt/kvm/kvm_main.c        |  2 --
>  6 files changed, 103 insertions(+), 60 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index 3a898712d6cd..4a039341dc29 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
>  	void __user *argp = (void __user *)arg;
>  	long r;
>  
> +	if (ioctl = KVM_INTERRUPT) {

I would add a comment here that this ioctl is async to vcpu execution,
so it is understandable why you skip the vcpu_load().

[As an aside, it is nice that this is now more obvious when looking at
the architectures' handlers.]

> +		struct kvm_mips_interrupt irq;
> +
> +		if (copy_from_user(&irq, argp, sizeof(irq)))
> +			return -EFAULT;
> +		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
> +			  irq.irq);
> +
> +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> +	}
> +
> +	vcpu_load(vcpu);
> +
>  	switch (ioctl) {

(...)

> diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
> index c06bc9552438..6b5dd3a25fe8 100644
> --- a/arch/powerpc/kvm/powerpc.c
> +++ b/arch/powerpc/kvm/powerpc.c
> @@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	void __user *argp = (void __user *)arg;
>  	long r;
>  
> -	switch (ioctl) {
> -	case KVM_INTERRUPT: {
> +	if (ioctl = KVM_INTERRUPT) {

Same here.

>  		struct kvm_interrupt irq;
> -		r = -EFAULT;
>  		if (copy_from_user(&irq, argp, sizeof(irq)))
> -			goto out;
> -		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> -		goto out;
> +			return -EFAULT;
> +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
>  	}
>  
> +	vcpu_load(vcpu);
> +
> +	switch (ioctl) {
>  	case KVM_ENABLE_CAP:
>  	{
>  		struct kvm_enable_cap cap;
> @@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	}
>  
>  out:
> +	vcpu_put(vcpu);
>  	return r;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 43278f334ce3..cd067b63d77f 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	case KVM_S390_IRQ: {
>  		struct kvm_s390_irq s390irq;
>  
> -		r = -EFAULT;
>  		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
> -			break;
> -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> -		break;
> +			return -EFAULT;
> +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
>  	}
>  	case KVM_S390_INTERRUPT: {
>  		struct kvm_s390_interrupt s390int;
>  		struct kvm_s390_irq s390irq;
>  
> -		r = -EFAULT;
>  		if (copy_from_user(&s390int, argp, sizeof(s390int)))
> -			break;
> +			return -EFAULT;
>  		if (s390int_to_s390irq(&s390int, &s390irq))
>  			return -EINVAL;
> -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> -		break;
> +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
>  	}
> +	}

I find the special casing with the immediate return a bit ugly. Maybe
introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
its default case and return here if ret != -ENOIOCTLCMD? Christian,
what do you think?

> +
> +	vcpu_load(vcpu);
> +
> +	switch (ioctl) {
>  	case KVM_S390_STORE_STATUS:
>  		idx = srcu_read_lock(&vcpu->kvm->srcu);
>  		r = kvm_s390_vcpu_store_status(vcpu, arg);
> @@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	default:
>  		r = -ENOTTY;
>  	}
> +
> +	vcpu_put(vcpu);
>  	return r;
>  }
>  

(...)

> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 06751bbecd58..ad5f83159a15 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		break;
>  	}
>  	default:
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
> -		vcpu_put(vcpu);
>  	}
>  out:
>  	mutex_unlock(&vcpu->mutex);

It would be nice if we could get rid of the special casing at the
beginning of this function, but as it would involve not taking the
mutex for special cases (and not releasing it for those special cases),
I don't see an elegant way to do that.

[PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Cornelia Huck]

On Mon,  4 Dec 2017 21:35:36 +0100
Christoffer Dall <cdall@kernel.org> wrote:

> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> Move the calls to vcpu_load() and vcpu_put() in to the architecture
> specific implementations of kvm_arch_vcpu_ioctl() which dispatches
> further architecture-specific ioctls on to other functions.
> 
> Some architectures support asynchronous vcpu ioctls which cannot call
> vcpu_load() or take the vcpu->mutex, because that would prevent
> concurrent execution with a running VCPU, which is the intended purpose
> of these ioctls, for example because they inject interrupts.
> 
> We repeat the separate checks for these specifics in the architecture
> code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
> calling vcpu_load for these ioctls.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> ---
>  arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
>  arch/powerpc/kvm/powerpc.c | 13 ++++++-----
>  arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
>  arch/x86/kvm/x86.c         | 22 +++++++++++++-----
>  virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
>  virt/kvm/kvm_main.c        |  2 --
>  6 files changed, 103 insertions(+), 60 deletions(-)
> 
> diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> index 3a898712d6cd..4a039341dc29 100644
> --- a/arch/mips/kvm/mips.c
> +++ b/arch/mips/kvm/mips.c
> @@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
>  	void __user *argp = (void __user *)arg;
>  	long r;
>  
> +	if (ioctl == KVM_INTERRUPT) {

I would add a comment here that this ioctl is async to vcpu execution,
so it is understandable why you skip the vcpu_load().

[As an aside, it is nice that this is now more obvious when looking at
the architectures' handlers.]

> +		struct kvm_mips_interrupt irq;
> +
> +		if (copy_from_user(&irq, argp, sizeof(irq)))
> +			return -EFAULT;
> +		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
> +			  irq.irq);
> +
> +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> +	}
> +
> +	vcpu_load(vcpu);
> +
>  	switch (ioctl) {

(...)

> diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
> index c06bc9552438..6b5dd3a25fe8 100644
> --- a/arch/powerpc/kvm/powerpc.c
> +++ b/arch/powerpc/kvm/powerpc.c
> @@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	void __user *argp = (void __user *)arg;
>  	long r;
>  
> -	switch (ioctl) {
> -	case KVM_INTERRUPT: {
> +	if (ioctl == KVM_INTERRUPT) {

Same here.

>  		struct kvm_interrupt irq;
> -		r = -EFAULT;
>  		if (copy_from_user(&irq, argp, sizeof(irq)))
> -			goto out;
> -		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> -		goto out;
> +			return -EFAULT;
> +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
>  	}
>  
> +	vcpu_load(vcpu);
> +
> +	switch (ioctl) {
>  	case KVM_ENABLE_CAP:
>  	{
>  		struct kvm_enable_cap cap;
> @@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	}
>  
>  out:
> +	vcpu_put(vcpu);
>  	return r;
>  }
>  
> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> index 43278f334ce3..cd067b63d77f 100644
> --- a/arch/s390/kvm/kvm-s390.c
> +++ b/arch/s390/kvm/kvm-s390.c
> @@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	case KVM_S390_IRQ: {
>  		struct kvm_s390_irq s390irq;
>  
> -		r = -EFAULT;
>  		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
> -			break;
> -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> -		break;
> +			return -EFAULT;
> +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
>  	}
>  	case KVM_S390_INTERRUPT: {
>  		struct kvm_s390_interrupt s390int;
>  		struct kvm_s390_irq s390irq;
>  
> -		r = -EFAULT;
>  		if (copy_from_user(&s390int, argp, sizeof(s390int)))
> -			break;
> +			return -EFAULT;
>  		if (s390int_to_s390irq(&s390int, &s390irq))
>  			return -EINVAL;
> -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> -		break;
> +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
>  	}
> +	}

I find the special casing with the immediate return a bit ugly. Maybe
introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
its default case and return here if ret != -ENOIOCTLCMD? Christian,
what do you think?

> +
> +	vcpu_load(vcpu);
> +
> +	switch (ioctl) {
>  	case KVM_S390_STORE_STATUS:
>  		idx = srcu_read_lock(&vcpu->kvm->srcu);
>  		r = kvm_s390_vcpu_store_status(vcpu, arg);
> @@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
>  	default:
>  		r = -ENOTTY;
>  	}
> +
> +	vcpu_put(vcpu);
>  	return r;
>  }
>  

(...)

> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 06751bbecd58..ad5f83159a15 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
>  		break;
>  	}
>  	default:
> -		vcpu_load(vcpu);
>  		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
> -		vcpu_put(vcpu);
>  	}
>  out:
>  	mutex_unlock(&vcpu->mutex);

It would be nice if we could get rid of the special casing at the
beginning of this function, but as it would involve not taking the
mutex for special cases (and not releasing it for those special cases),
I don't see an elegant way to do that.

Re: [PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Christoffer Dall]

On Mon, Dec 11, 2017 at 02:12:41PM +0100, Cornelia Huck wrote:
> On Mon,  4 Dec 2017 21:35:36 +0100
> Christoffer Dall <cdall@kernel.org> wrote:
> 
> > From: Christoffer Dall <christoffer.dall@linaro.org>
> > 
> > Move the calls to vcpu_load() and vcpu_put() in to the architecture
> > specific implementations of kvm_arch_vcpu_ioctl() which dispatches
> > further architecture-specific ioctls on to other functions.
> > 
> > Some architectures support asynchronous vcpu ioctls which cannot call
> > vcpu_load() or take the vcpu->mutex, because that would prevent
> > concurrent execution with a running VCPU, which is the intended purpose
> > of these ioctls, for example because they inject interrupts.
> > 
> > We repeat the separate checks for these specifics in the architecture
> > code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
> > calling vcpu_load for these ioctls.
> > 
> > Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> > ---
> >  arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
> >  arch/powerpc/kvm/powerpc.c | 13 ++++++-----
> >  arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
> >  arch/x86/kvm/x86.c         | 22 +++++++++++++-----
> >  virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
> >  virt/kvm/kvm_main.c        |  2 --
> >  6 files changed, 103 insertions(+), 60 deletions(-)
> > 
> > diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> > index 3a898712d6cd..4a039341dc29 100644
> > --- a/arch/mips/kvm/mips.c
> > +++ b/arch/mips/kvm/mips.c
> > @@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
> >  	void __user *argp = (void __user *)arg;
> >  	long r;
> >  
> > +	if (ioctl == KVM_INTERRUPT) {
> 
> I would add a comment here that this ioctl is async to vcpu execution,
> so it is understandable why you skip the vcpu_load().

Yes, that would be appropriate.

> 
> [As an aside, it is nice that this is now more obvious when looking at
> the architectures' handlers.]
> 

Agreed.

> > +		struct kvm_mips_interrupt irq;
> > +
> > +		if (copy_from_user(&irq, argp, sizeof(irq)))
> > +			return -EFAULT;
> > +		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
> > +			  irq.irq);
> > +
> > +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> > +	}
> > +
> > +	vcpu_load(vcpu);
> > +
> >  	switch (ioctl) {
> 
> (...)
> 
> > diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
> > index c06bc9552438..6b5dd3a25fe8 100644
> > --- a/arch/powerpc/kvm/powerpc.c
> > +++ b/arch/powerpc/kvm/powerpc.c
> > @@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	void __user *argp = (void __user *)arg;
> >  	long r;
> >  
> > -	switch (ioctl) {
> > -	case KVM_INTERRUPT: {
> > +	if (ioctl == KVM_INTERRUPT) {
> 
> Same here.
> 
> >  		struct kvm_interrupt irq;
> > -		r = -EFAULT;
> >  		if (copy_from_user(&irq, argp, sizeof(irq)))
> > -			goto out;
> > -		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> > -		goto out;
> > +			return -EFAULT;
> > +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> >  	}
> >  
> > +	vcpu_load(vcpu);
> > +
> > +	switch (ioctl) {
> >  	case KVM_ENABLE_CAP:
> >  	{
> >  		struct kvm_enable_cap cap;
> > @@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	}
> >  
> >  out:
> > +	vcpu_put(vcpu);
> >  	return r;
> >  }
> >  
> > diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> > index 43278f334ce3..cd067b63d77f 100644
> > --- a/arch/s390/kvm/kvm-s390.c
> > +++ b/arch/s390/kvm/kvm-s390.c
> > @@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	case KVM_S390_IRQ: {
> >  		struct kvm_s390_irq s390irq;
> >  
> > -		r = -EFAULT;
> >  		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
> > -			break;
> > -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> > -		break;
> > +			return -EFAULT;
> > +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
> >  	}
> >  	case KVM_S390_INTERRUPT: {
> >  		struct kvm_s390_interrupt s390int;
> >  		struct kvm_s390_irq s390irq;
> >  
> > -		r = -EFAULT;
> >  		if (copy_from_user(&s390int, argp, sizeof(s390int)))
> > -			break;
> > +			return -EFAULT;
> >  		if (s390int_to_s390irq(&s390int, &s390irq))
> >  			return -EINVAL;
> > -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> > -		break;
> > +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
> >  	}
> > +	}
> 
> I find the special casing with the immediate return a bit ugly. Maybe
> introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
> its default case and return here if ret != -ENOIOCTLCMD? Christian,
> what do you think?
> 
> > +
> > +	vcpu_load(vcpu);
> > +
> > +	switch (ioctl) {
> >  	case KVM_S390_STORE_STATUS:
> >  		idx = srcu_read_lock(&vcpu->kvm->srcu);
> >  		r = kvm_s390_vcpu_store_status(vcpu, arg);
> > @@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	default:
> >  		r = -ENOTTY;
> >  	}
> > +
> > +	vcpu_put(vcpu);
> >  	return r;
> >  }
> >  
> 
> (...)
> 
> > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> > index 06751bbecd58..ad5f83159a15 100644
> > --- a/virt/kvm/kvm_main.c
> > +++ b/virt/kvm/kvm_main.c
> > @@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
> >  		break;
> >  	}
> >  	default:
> > -		vcpu_load(vcpu);
> >  		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
> > -		vcpu_put(vcpu);
> >  	}
> >  out:
> >  	mutex_unlock(&vcpu->mutex);
> 
> It would be nice if we could get rid of the special casing at the
> beginning of this function, but as it would involve not taking the
> mutex for special cases (and not releasing it for those special cases),
> I don't see an elegant way to do that.

I would also have liked that, and that's essentially what I had in the
first version, but Paolo thought the result was too high an increase in
complexity in the architecture-specfic functions throughout.  I don't
have any better suggestions either.

Thanks,
-Christoffer

Re: [PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Christoffer Dall]

On Mon, Dec 11, 2017 at 02:12:41PM +0100, Cornelia Huck wrote:
> On Mon,  4 Dec 2017 21:35:36 +0100
> Christoffer Dall <cdall@kernel.org> wrote:
> 
> > From: Christoffer Dall <christoffer.dall@linaro.org>
> > 
> > Move the calls to vcpu_load() and vcpu_put() in to the architecture
> > specific implementations of kvm_arch_vcpu_ioctl() which dispatches
> > further architecture-specific ioctls on to other functions.
> > 
> > Some architectures support asynchronous vcpu ioctls which cannot call
> > vcpu_load() or take the vcpu->mutex, because that would prevent
> > concurrent execution with a running VCPU, which is the intended purpose
> > of these ioctls, for example because they inject interrupts.
> > 
> > We repeat the separate checks for these specifics in the architecture
> > code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
> > calling vcpu_load for these ioctls.
> > 
> > Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> > ---
> >  arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
> >  arch/powerpc/kvm/powerpc.c | 13 ++++++-----
> >  arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
> >  arch/x86/kvm/x86.c         | 22 +++++++++++++-----
> >  virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
> >  virt/kvm/kvm_main.c        |  2 --
> >  6 files changed, 103 insertions(+), 60 deletions(-)
> > 
> > diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> > index 3a898712d6cd..4a039341dc29 100644
> > --- a/arch/mips/kvm/mips.c
> > +++ b/arch/mips/kvm/mips.c
> > @@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
> >  	void __user *argp = (void __user *)arg;
> >  	long r;
> >  
> > +	if (ioctl = KVM_INTERRUPT) {
> 
> I would add a comment here that this ioctl is async to vcpu execution,
> so it is understandable why you skip the vcpu_load().

Yes, that would be appropriate.

> 
> [As an aside, it is nice that this is now more obvious when looking at
> the architectures' handlers.]
> 

Agreed.

> > +		struct kvm_mips_interrupt irq;
> > +
> > +		if (copy_from_user(&irq, argp, sizeof(irq)))
> > +			return -EFAULT;
> > +		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
> > +			  irq.irq);
> > +
> > +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> > +	}
> > +
> > +	vcpu_load(vcpu);
> > +
> >  	switch (ioctl) {
> 
> (...)
> 
> > diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
> > index c06bc9552438..6b5dd3a25fe8 100644
> > --- a/arch/powerpc/kvm/powerpc.c
> > +++ b/arch/powerpc/kvm/powerpc.c
> > @@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	void __user *argp = (void __user *)arg;
> >  	long r;
> >  
> > -	switch (ioctl) {
> > -	case KVM_INTERRUPT: {
> > +	if (ioctl = KVM_INTERRUPT) {
> 
> Same here.
> 
> >  		struct kvm_interrupt irq;
> > -		r = -EFAULT;
> >  		if (copy_from_user(&irq, argp, sizeof(irq)))
> > -			goto out;
> > -		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> > -		goto out;
> > +			return -EFAULT;
> > +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> >  	}
> >  
> > +	vcpu_load(vcpu);
> > +
> > +	switch (ioctl) {
> >  	case KVM_ENABLE_CAP:
> >  	{
> >  		struct kvm_enable_cap cap;
> > @@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	}
> >  
> >  out:
> > +	vcpu_put(vcpu);
> >  	return r;
> >  }
> >  
> > diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> > index 43278f334ce3..cd067b63d77f 100644
> > --- a/arch/s390/kvm/kvm-s390.c
> > +++ b/arch/s390/kvm/kvm-s390.c
> > @@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	case KVM_S390_IRQ: {
> >  		struct kvm_s390_irq s390irq;
> >  
> > -		r = -EFAULT;
> >  		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
> > -			break;
> > -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> > -		break;
> > +			return -EFAULT;
> > +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
> >  	}
> >  	case KVM_S390_INTERRUPT: {
> >  		struct kvm_s390_interrupt s390int;
> >  		struct kvm_s390_irq s390irq;
> >  
> > -		r = -EFAULT;
> >  		if (copy_from_user(&s390int, argp, sizeof(s390int)))
> > -			break;
> > +			return -EFAULT;
> >  		if (s390int_to_s390irq(&s390int, &s390irq))
> >  			return -EINVAL;
> > -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> > -		break;
> > +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
> >  	}
> > +	}
> 
> I find the special casing with the immediate return a bit ugly. Maybe
> introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
> its default case and return here if ret != -ENOIOCTLCMD? Christian,
> what do you think?
> 
> > +
> > +	vcpu_load(vcpu);
> > +
> > +	switch (ioctl) {
> >  	case KVM_S390_STORE_STATUS:
> >  		idx = srcu_read_lock(&vcpu->kvm->srcu);
> >  		r = kvm_s390_vcpu_store_status(vcpu, arg);
> > @@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	default:
> >  		r = -ENOTTY;
> >  	}
> > +
> > +	vcpu_put(vcpu);
> >  	return r;
> >  }
> >  
> 
> (...)
> 
> > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> > index 06751bbecd58..ad5f83159a15 100644
> > --- a/virt/kvm/kvm_main.c
> > +++ b/virt/kvm/kvm_main.c
> > @@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
> >  		break;
> >  	}
> >  	default:
> > -		vcpu_load(vcpu);
> >  		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
> > -		vcpu_put(vcpu);
> >  	}
> >  out:
> >  	mutex_unlock(&vcpu->mutex);
> 
> It would be nice if we could get rid of the special casing at the
> beginning of this function, but as it would involve not taking the
> mutex for special cases (and not releasing it for those special cases),
> I don't see an elegant way to do that.

I would also have liked that, and that's essentially what I had in the
first version, but Paolo thought the result was too high an increase in
complexity in the architecture-specfic functions throughout.  I don't
have any better suggestions either.

Thanks,
-Christoffer

[PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Christoffer Dall]

On Mon, Dec 11, 2017 at 02:12:41PM +0100, Cornelia Huck wrote:
> On Mon,  4 Dec 2017 21:35:36 +0100
> Christoffer Dall <cdall@kernel.org> wrote:
> 
> > From: Christoffer Dall <christoffer.dall@linaro.org>
> > 
> > Move the calls to vcpu_load() and vcpu_put() in to the architecture
> > specific implementations of kvm_arch_vcpu_ioctl() which dispatches
> > further architecture-specific ioctls on to other functions.
> > 
> > Some architectures support asynchronous vcpu ioctls which cannot call
> > vcpu_load() or take the vcpu->mutex, because that would prevent
> > concurrent execution with a running VCPU, which is the intended purpose
> > of these ioctls, for example because they inject interrupts.
> > 
> > We repeat the separate checks for these specifics in the architecture
> > code for MIPS, S390 and PPC, and avoid taking the vcpu->mutex and
> > calling vcpu_load for these ioctls.
> > 
> > Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> > ---
> >  arch/mips/kvm/mips.c       | 49 +++++++++++++++++++++++----------------
> >  arch/powerpc/kvm/powerpc.c | 13 ++++++-----
> >  arch/s390/kvm/kvm-s390.c   | 19 ++++++++-------
> >  arch/x86/kvm/x86.c         | 22 +++++++++++++-----
> >  virt/kvm/arm/arm.c         | 58 ++++++++++++++++++++++++++++++++--------------
> >  virt/kvm/kvm_main.c        |  2 --
> >  6 files changed, 103 insertions(+), 60 deletions(-)
> > 
> > diff --git a/arch/mips/kvm/mips.c b/arch/mips/kvm/mips.c
> > index 3a898712d6cd..4a039341dc29 100644
> > --- a/arch/mips/kvm/mips.c
> > +++ b/arch/mips/kvm/mips.c
> > @@ -913,56 +913,65 @@ long kvm_arch_vcpu_ioctl(struct file *filp, unsigned int ioctl,
> >  	void __user *argp = (void __user *)arg;
> >  	long r;
> >  
> > +	if (ioctl == KVM_INTERRUPT) {
> 
> I would add a comment here that this ioctl is async to vcpu execution,
> so it is understandable why you skip the vcpu_load().

Yes, that would be appropriate.

> 
> [As an aside, it is nice that this is now more obvious when looking at
> the architectures' handlers.]
> 

Agreed.

> > +		struct kvm_mips_interrupt irq;
> > +
> > +		if (copy_from_user(&irq, argp, sizeof(irq)))
> > +			return -EFAULT;
> > +		kvm_debug("[%d] %s: irq: %d\n", vcpu->vcpu_id, __func__,
> > +			  irq.irq);
> > +
> > +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> > +	}
> > +
> > +	vcpu_load(vcpu);
> > +
> >  	switch (ioctl) {
> 
> (...)
> 
> > diff --git a/arch/powerpc/kvm/powerpc.c b/arch/powerpc/kvm/powerpc.c
> > index c06bc9552438..6b5dd3a25fe8 100644
> > --- a/arch/powerpc/kvm/powerpc.c
> > +++ b/arch/powerpc/kvm/powerpc.c
> > @@ -1617,16 +1617,16 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	void __user *argp = (void __user *)arg;
> >  	long r;
> >  
> > -	switch (ioctl) {
> > -	case KVM_INTERRUPT: {
> > +	if (ioctl == KVM_INTERRUPT) {
> 
> Same here.
> 
> >  		struct kvm_interrupt irq;
> > -		r = -EFAULT;
> >  		if (copy_from_user(&irq, argp, sizeof(irq)))
> > -			goto out;
> > -		r = kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> > -		goto out;
> > +			return -EFAULT;
> > +		return kvm_vcpu_ioctl_interrupt(vcpu, &irq);
> >  	}
> >  
> > +	vcpu_load(vcpu);
> > +
> > +	switch (ioctl) {
> >  	case KVM_ENABLE_CAP:
> >  	{
> >  		struct kvm_enable_cap cap;
> > @@ -1666,6 +1666,7 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	}
> >  
> >  out:
> > +	vcpu_put(vcpu);
> >  	return r;
> >  }
> >  
> > diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
> > index 43278f334ce3..cd067b63d77f 100644
> > --- a/arch/s390/kvm/kvm-s390.c
> > +++ b/arch/s390/kvm/kvm-s390.c
> > @@ -3743,24 +3743,25 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	case KVM_S390_IRQ: {
> >  		struct kvm_s390_irq s390irq;
> >  
> > -		r = -EFAULT;
> >  		if (copy_from_user(&s390irq, argp, sizeof(s390irq)))
> > -			break;
> > -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> > -		break;
> > +			return -EFAULT;
> > +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
> >  	}
> >  	case KVM_S390_INTERRUPT: {
> >  		struct kvm_s390_interrupt s390int;
> >  		struct kvm_s390_irq s390irq;
> >  
> > -		r = -EFAULT;
> >  		if (copy_from_user(&s390int, argp, sizeof(s390int)))
> > -			break;
> > +			return -EFAULT;
> >  		if (s390int_to_s390irq(&s390int, &s390irq))
> >  			return -EINVAL;
> > -		r = kvm_s390_inject_vcpu(vcpu, &s390irq);
> > -		break;
> > +		return kvm_s390_inject_vcpu(vcpu, &s390irq);
> >  	}
> > +	}
> 
> I find the special casing with the immediate return a bit ugly. Maybe
> introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
> its default case and return here if ret != -ENOIOCTLCMD? Christian,
> what do you think?
> 
> > +
> > +	vcpu_load(vcpu);
> > +
> > +	switch (ioctl) {
> >  	case KVM_S390_STORE_STATUS:
> >  		idx = srcu_read_lock(&vcpu->kvm->srcu);
> >  		r = kvm_s390_vcpu_store_status(vcpu, arg);
> > @@ -3883,6 +3884,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
> >  	default:
> >  		r = -ENOTTY;
> >  	}
> > +
> > +	vcpu_put(vcpu);
> >  	return r;
> >  }
> >  
> 
> (...)
> 
> > diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> > index 06751bbecd58..ad5f83159a15 100644
> > --- a/virt/kvm/kvm_main.c
> > +++ b/virt/kvm/kvm_main.c
> > @@ -2693,9 +2693,7 @@ static long kvm_vcpu_ioctl(struct file *filp,
> >  		break;
> >  	}
> >  	default:
> > -		vcpu_load(vcpu);
> >  		r = kvm_arch_vcpu_ioctl(filp, ioctl, arg);
> > -		vcpu_put(vcpu);
> >  	}
> >  out:
> >  	mutex_unlock(&vcpu->mutex);
> 
> It would be nice if we could get rid of the special casing at the
> beginning of this function, but as it would involve not taking the
> mutex for special cases (and not releasing it for those special cases),
> I don't see an elegant way to do that.

I would also have liked that, and that's essentially what I had in the
first version, but Paolo thought the result was too high an increase in
complexity in the architecture-specfic functions throughout.  I don't
have any better suggestions either.

Thanks,
-Christoffer

Re: [PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Paolo Bonzini]

On 11/12/2017 16:22, Christoffer Dall wrote:
> I find the special casing with the immediate return a bit ugly. Maybe
> introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
> its default case and return here if ret != -ENOIOCTLCMD? Christian,
> what do you think?

I'll post my attempt at it shortly.

Thanks,

Paolo

Re: [PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Paolo Bonzini]

On 11/12/2017 16:22, Christoffer Dall wrote:
> I find the special casing with the immediate return a bit ugly. Maybe
> introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
> its default case and return here if ret != -ENOIOCTLCMD? Christian,
> what do you think?

I'll post my attempt at it shortly.

Thanks,

Paolo

[PATCH v3 14/16] KVM: Move vcpu_load to arch-specific kvm_arch_vcpu_ioctl

[Paolo Bonzini]

On 11/12/2017 16:22, Christoffer Dall wrote:
> I find the special casing with the immediate return a bit ugly. Maybe
> introduce a helper async_vcpu_ioctl() or so that sets -ENOIOCTLCMD in
> its default case and return here if ret != -ENOIOCTLCMD? Christian,
> what do you think?

I'll post my attempt at it shortly.

Thanks,

Paolo

[PATCH v3 15/16] KVM: arm/arm64: Avoid vcpu_load for other vcpu ioctls than KVM_RUN

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Calling vcpu_load() registers preempt notifiers for this vcpu and calls
kvm_arch_vcpu_load().  The latter will soon be doing a lot of heavy
lifting on arm/arm64 and will try to do things such as enabling the
virtual timer and setting us up to handle interrupts from the timer
hardware.

Loading state onto hardware registers and enabling hardware to signal
interrupts can be problematic when we're not actually about to run the
VCPU, because it makes it difficult to establish the right context when
handling interrupts from the timer, and it makes the register access
code difficult to reason about.

Luckily, now when we call vcpu_load in each ioctl implementation, we can
simply remove the call from the non-KVM_RUN vcpu ioctls, and our
kvm_arch_vcpu_load() is only used for loading vcpu content to the
physical CPU when we're actually going to run the vcpu.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm64/kvm/guest.c | 3 ---
 virt/kvm/arm/arm.c     | 9 ---------
 2 files changed, 12 deletions(-)

diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index d7e3299a7734..959e50d2588c 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -363,8 +363,6 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	int ret = 0;
 
-	vcpu_load(vcpu);
-
 	trace_kvm_set_guest_debug(vcpu, dbg->control);
 
 	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK) {
@@ -386,7 +384,6 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	}
 
 out:
-	vcpu_put(vcpu);
 	return ret;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 8223c59be507..a760ef1803be 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -381,14 +381,11 @@ static void vcpu_power_off(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
-	vcpu_load(vcpu);
-
 	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 
-	vcpu_put(vcpu);
 	return 0;
 }
 
@@ -397,8 +394,6 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	int ret = 0;
 
-	vcpu_load(vcpu);
-
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
 		vcpu->arch.power_off = false;
@@ -410,7 +405,6 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		ret = -EINVAL;
 	}
 
-	vcpu_put(vcpu);
 	return ret;
 }
 
@@ -1003,8 +997,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	struct kvm_device_attr attr;
 	long r;
 
-	vcpu_load(vcpu);
-
 	switch (ioctl) {
 	case KVM_ARM_VCPU_INIT: {
 		struct kvm_vcpu_init init;
@@ -1081,7 +1073,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 	}
 
-	vcpu_put(vcpu);
 	return r;
 }
 
-- 
2.14.2

[PATCH v3 15/16] KVM: arm/arm64: Avoid vcpu_load for other vcpu ioctls than KVM_RUN

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Calling vcpu_load() registers preempt notifiers for this vcpu and calls
kvm_arch_vcpu_load().  The latter will soon be doing a lot of heavy
lifting on arm/arm64 and will try to do things such as enabling the
virtual timer and setting us up to handle interrupts from the timer
hardware.

Loading state onto hardware registers and enabling hardware to signal
interrupts can be problematic when we're not actually about to run the
VCPU, because it makes it difficult to establish the right context when
handling interrupts from the timer, and it makes the register access
code difficult to reason about.

Luckily, now when we call vcpu_load in each ioctl implementation, we can
simply remove the call from the non-KVM_RUN vcpu ioctls, and our
kvm_arch_vcpu_load() is only used for loading vcpu content to the
physical CPU when we're actually going to run the vcpu.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm64/kvm/guest.c | 3 ---
 virt/kvm/arm/arm.c     | 9 ---------
 2 files changed, 12 deletions(-)

diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index d7e3299a7734..959e50d2588c 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -363,8 +363,6 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	int ret = 0;
 
-	vcpu_load(vcpu);
-
 	trace_kvm_set_guest_debug(vcpu, dbg->control);
 
 	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK) {
@@ -386,7 +384,6 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	}
 
 out:
-	vcpu_put(vcpu);
 	return ret;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 8223c59be507..a760ef1803be 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -381,14 +381,11 @@ static void vcpu_power_off(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
-	vcpu_load(vcpu);
-
 	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 
-	vcpu_put(vcpu);
 	return 0;
 }
 
@@ -397,8 +394,6 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	int ret = 0;
 
-	vcpu_load(vcpu);
-
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
 		vcpu->arch.power_off = false;
@@ -410,7 +405,6 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		ret = -EINVAL;
 	}
 
-	vcpu_put(vcpu);
 	return ret;
 }
 
@@ -1003,8 +997,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	struct kvm_device_attr attr;
 	long r;
 
-	vcpu_load(vcpu);
-
 	switch (ioctl) {
 	case KVM_ARM_VCPU_INIT: {
 		struct kvm_vcpu_init init;
@@ -1081,7 +1073,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 	}
 
-	vcpu_put(vcpu);
 	return r;
 }
 
-- 
2.14.2

[PATCH v3 15/16] KVM: arm/arm64: Avoid vcpu_load for other vcpu ioctls than KVM_RUN

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Calling vcpu_load() registers preempt notifiers for this vcpu and calls
kvm_arch_vcpu_load().  The latter will soon be doing a lot of heavy
lifting on arm/arm64 and will try to do things such as enabling the
virtual timer and setting us up to handle interrupts from the timer
hardware.

Loading state onto hardware registers and enabling hardware to signal
interrupts can be problematic when we're not actually about to run the
VCPU, because it makes it difficult to establish the right context when
handling interrupts from the timer, and it makes the register access
code difficult to reason about.

Luckily, now when we call vcpu_load in each ioctl implementation, we can
simply remove the call from the non-KVM_RUN vcpu ioctls, and our
kvm_arch_vcpu_load() is only used for loading vcpu content to the
physical CPU when we're actually going to run the vcpu.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm64/kvm/guest.c | 3 ---
 virt/kvm/arm/arm.c     | 9 ---------
 2 files changed, 12 deletions(-)

diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index d7e3299a7734..959e50d2588c 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -363,8 +363,6 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	int ret = 0;
 
-	vcpu_load(vcpu);
-
 	trace_kvm_set_guest_debug(vcpu, dbg->control);
 
 	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK) {
@@ -386,7 +384,6 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	}
 
 out:
-	vcpu_put(vcpu);
 	return ret;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 8223c59be507..a760ef1803be 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -381,14 +381,11 @@ static void vcpu_power_off(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
-	vcpu_load(vcpu);
-
 	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 
-	vcpu_put(vcpu);
 	return 0;
 }
 
@@ -397,8 +394,6 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	int ret = 0;
 
-	vcpu_load(vcpu);
-
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
 		vcpu->arch.power_off = false;
@@ -410,7 +405,6 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		ret = -EINVAL;
 	}
 
-	vcpu_put(vcpu);
 	return ret;
 }
 
@@ -1003,8 +997,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	struct kvm_device_attr attr;
 	long r;
 
-	vcpu_load(vcpu);
-
 	switch (ioctl) {
 	case KVM_ARM_VCPU_INIT: {
 		struct kvm_vcpu_init init;
@@ -1081,7 +1073,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 	}
 
-	vcpu_put(vcpu);
 	return r;
 }
 
-- 
2.14.2

[PATCH v3 15/16] KVM: arm/arm64: Avoid vcpu_load for other vcpu ioctls than KVM_RUN

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Calling vcpu_load() registers preempt notifiers for this vcpu and calls
kvm_arch_vcpu_load().  The latter will soon be doing a lot of heavy
lifting on arm/arm64 and will try to do things such as enabling the
virtual timer and setting us up to handle interrupts from the timer
hardware.

Loading state onto hardware registers and enabling hardware to signal
interrupts can be problematic when we're not actually about to run the
VCPU, because it makes it difficult to establish the right context when
handling interrupts from the timer, and it makes the register access
code difficult to reason about.

Luckily, now when we call vcpu_load in each ioctl implementation, we can
simply remove the call from the non-KVM_RUN vcpu ioctls, and our
kvm_arch_vcpu_load() is only used for loading vcpu content to the
physical CPU when we're actually going to run the vcpu.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 arch/arm64/kvm/guest.c | 3 ---
 virt/kvm/arm/arm.c     | 9 ---------
 2 files changed, 12 deletions(-)

diff --git a/arch/arm64/kvm/guest.c b/arch/arm64/kvm/guest.c
index d7e3299a7734..959e50d2588c 100644
--- a/arch/arm64/kvm/guest.c
+++ b/arch/arm64/kvm/guest.c
@@ -363,8 +363,6 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 {
 	int ret = 0;
 
-	vcpu_load(vcpu);
-
 	trace_kvm_set_guest_debug(vcpu, dbg->control);
 
 	if (dbg->control & ~KVM_GUESTDBG_VALID_MASK) {
@@ -386,7 +384,6 @@ int kvm_arch_vcpu_ioctl_set_guest_debug(struct kvm_vcpu *vcpu,
 	}
 
 out:
-	vcpu_put(vcpu);
 	return ret;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index 8223c59be507..a760ef1803be 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -381,14 +381,11 @@ static void vcpu_power_off(struct kvm_vcpu *vcpu)
 int kvm_arch_vcpu_ioctl_get_mpstate(struct kvm_vcpu *vcpu,
 				    struct kvm_mp_state *mp_state)
 {
-	vcpu_load(vcpu);
-
 	if (vcpu->arch.power_off)
 		mp_state->mp_state = KVM_MP_STATE_STOPPED;
 	else
 		mp_state->mp_state = KVM_MP_STATE_RUNNABLE;
 
-	vcpu_put(vcpu);
 	return 0;
 }
 
@@ -397,8 +394,6 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 {
 	int ret = 0;
 
-	vcpu_load(vcpu);
-
 	switch (mp_state->mp_state) {
 	case KVM_MP_STATE_RUNNABLE:
 		vcpu->arch.power_off = false;
@@ -410,7 +405,6 @@ int kvm_arch_vcpu_ioctl_set_mpstate(struct kvm_vcpu *vcpu,
 		ret = -EINVAL;
 	}
 
-	vcpu_put(vcpu);
 	return ret;
 }
 
@@ -1003,8 +997,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 	struct kvm_device_attr attr;
 	long r;
 
-	vcpu_load(vcpu);
-
 	switch (ioctl) {
 	case KVM_ARM_VCPU_INIT: {
 		struct kvm_vcpu_init init;
@@ -1081,7 +1073,6 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		r = -EINVAL;
 	}
 
-	vcpu_put(vcpu);
 	return r;
 }
 
-- 
2.14.2

[PATCH v3 16/16] KVM: arm/arm64: Move vcpu_load call after kvm_vcpu_first_run_init

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Moving the call to vcpu_load() in kvm_arch_vcpu_ioctl_run() to after
we've called kvm_vcpu_first_run_init() simplifies some of the vgic and
there is also no need to do vcpu_load() for things such as handling the
immediate_exit flag.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 virt/kvm/arm/arch_timer.c     |  4 ----
 virt/kvm/arm/arm.c            | 12 +++++-------
 virt/kvm/arm/vgic/vgic-init.c | 11 -----------
 3 files changed, 5 insertions(+), 22 deletions(-)

diff --git a/virt/kvm/arm/arch_timer.c b/virt/kvm/arm/arch_timer.c
index 4151250ce8da..801fecfee299 100644
--- a/virt/kvm/arm/arch_timer.c
+++ b/virt/kvm/arm/arch_timer.c
@@ -839,11 +839,7 @@ int kvm_timer_enable(struct kvm_vcpu *vcpu)
 		return ret;
 
 no_vgic:
-	preempt_disable();
 	timer->enabled = 1;
-	kvm_timer_vcpu_load_vgic(vcpu);
-	preempt_enable();
-
 	return 0;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a760ef1803be..991f1aa70fb9 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -622,8 +622,6 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (unlikely(!kvm_vcpu_initialized(vcpu)))
 		return -ENOEXEC;
 
-	vcpu_load(vcpu);
-
 	ret = kvm_vcpu_first_run_init(vcpu);
 	if (ret)
 		goto out;
@@ -631,13 +629,13 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (run->exit_reason == KVM_EXIT_MMIO) {
 		ret = kvm_handle_mmio_return(vcpu, vcpu->run);
 		if (ret)
-			goto out;
+			return ret;
 	}
 
-	if (run->immediate_exit) {
-		ret = -EINTR;
-		goto out;
-	}
+	if (run->immediate_exit)
+		return -EINTR;
+
+	vcpu_load(vcpu);
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index 62310122ee78..a0688ef52ad7 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -300,17 +300,6 @@ int vgic_init(struct kvm *kvm)
 
 	dist->initialized = true;
 
-	/*
-	 * If we're initializing GICv2 on-demand when first running the VCPU
-	 * then we need to load the VGIC state onto the CPU.  We can detect
-	 * this easily by checking if we are in between vcpu_load and vcpu_put
-	 * when we just initialized the VGIC.
-	 */
-	preempt_disable();
-	vcpu = kvm_arm_get_running_vcpu();
-	if (vcpu)
-		kvm_vgic_load(vcpu);
-	preempt_enable();
 out:
 	return ret;
 }
-- 
2.14.2

[PATCH v3 16/16] KVM: arm/arm64: Move vcpu_load call after kvm_vcpu_first_run_init

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Moving the call to vcpu_load() in kvm_arch_vcpu_ioctl_run() to after
we've called kvm_vcpu_first_run_init() simplifies some of the vgic and
there is also no need to do vcpu_load() for things such as handling the
immediate_exit flag.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 virt/kvm/arm/arch_timer.c     |  4 ----
 virt/kvm/arm/arm.c            | 12 +++++-------
 virt/kvm/arm/vgic/vgic-init.c | 11 -----------
 3 files changed, 5 insertions(+), 22 deletions(-)

diff --git a/virt/kvm/arm/arch_timer.c b/virt/kvm/arm/arch_timer.c
index 4151250ce8da..801fecfee299 100644
--- a/virt/kvm/arm/arch_timer.c
+++ b/virt/kvm/arm/arch_timer.c
@@ -839,11 +839,7 @@ int kvm_timer_enable(struct kvm_vcpu *vcpu)
 		return ret;
 
 no_vgic:
-	preempt_disable();
 	timer->enabled = 1;
-	kvm_timer_vcpu_load_vgic(vcpu);
-	preempt_enable();
-
 	return 0;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a760ef1803be..991f1aa70fb9 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -622,8 +622,6 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (unlikely(!kvm_vcpu_initialized(vcpu)))
 		return -ENOEXEC;
 
-	vcpu_load(vcpu);
-
 	ret = kvm_vcpu_first_run_init(vcpu);
 	if (ret)
 		goto out;
@@ -631,13 +629,13 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (run->exit_reason = KVM_EXIT_MMIO) {
 		ret = kvm_handle_mmio_return(vcpu, vcpu->run);
 		if (ret)
-			goto out;
+			return ret;
 	}
 
-	if (run->immediate_exit) {
-		ret = -EINTR;
-		goto out;
-	}
+	if (run->immediate_exit)
+		return -EINTR;
+
+	vcpu_load(vcpu);
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index 62310122ee78..a0688ef52ad7 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -300,17 +300,6 @@ int vgic_init(struct kvm *kvm)
 
 	dist->initialized = true;
 
-	/*
-	 * If we're initializing GICv2 on-demand when first running the VCPU
-	 * then we need to load the VGIC state onto the CPU.  We can detect
-	 * this easily by checking if we are in between vcpu_load and vcpu_put
-	 * when we just initialized the VGIC.
-	 */
-	preempt_disable();
-	vcpu = kvm_arm_get_running_vcpu();
-	if (vcpu)
-		kvm_vgic_load(vcpu);
-	preempt_enable();
 out:
 	return ret;
 }
-- 
2.14.2

[PATCH v3 16/16] KVM: arm/arm64: Move vcpu_load call after kvm_vcpu_first_run_init

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Moving the call to vcpu_load() in kvm_arch_vcpu_ioctl_run() to after
we've called kvm_vcpu_first_run_init() simplifies some of the vgic and
there is also no need to do vcpu_load() for things such as handling the
immediate_exit flag.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 virt/kvm/arm/arch_timer.c     |  4 ----
 virt/kvm/arm/arm.c            | 12 +++++-------
 virt/kvm/arm/vgic/vgic-init.c | 11 -----------
 3 files changed, 5 insertions(+), 22 deletions(-)

diff --git a/virt/kvm/arm/arch_timer.c b/virt/kvm/arm/arch_timer.c
index 4151250ce8da..801fecfee299 100644
--- a/virt/kvm/arm/arch_timer.c
+++ b/virt/kvm/arm/arch_timer.c
@@ -839,11 +839,7 @@ int kvm_timer_enable(struct kvm_vcpu *vcpu)
 		return ret;
 
 no_vgic:
-	preempt_disable();
 	timer->enabled = 1;
-	kvm_timer_vcpu_load_vgic(vcpu);
-	preempt_enable();
-
 	return 0;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a760ef1803be..991f1aa70fb9 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -622,8 +622,6 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (unlikely(!kvm_vcpu_initialized(vcpu)))
 		return -ENOEXEC;
 
-	vcpu_load(vcpu);
-
 	ret = kvm_vcpu_first_run_init(vcpu);
 	if (ret)
 		goto out;
@@ -631,13 +629,13 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (run->exit_reason == KVM_EXIT_MMIO) {
 		ret = kvm_handle_mmio_return(vcpu, vcpu->run);
 		if (ret)
-			goto out;
+			return ret;
 	}
 
-	if (run->immediate_exit) {
-		ret = -EINTR;
-		goto out;
-	}
+	if (run->immediate_exit)
+		return -EINTR;
+
+	vcpu_load(vcpu);
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index 62310122ee78..a0688ef52ad7 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -300,17 +300,6 @@ int vgic_init(struct kvm *kvm)
 
 	dist->initialized = true;
 
-	/*
-	 * If we're initializing GICv2 on-demand when first running the VCPU
-	 * then we need to load the VGIC state onto the CPU.  We can detect
-	 * this easily by checking if we are in between vcpu_load and vcpu_put
-	 * when we just initialized the VGIC.
-	 */
-	preempt_disable();
-	vcpu = kvm_arm_get_running_vcpu();
-	if (vcpu)
-		kvm_vgic_load(vcpu);
-	preempt_enable();
 out:
 	return ret;
 }
-- 
2.14.2

[PATCH v3 16/16] KVM: arm/arm64: Move vcpu_load call after kvm_vcpu_first_run_init

[Christoffer Dall]

From: Christoffer Dall <christoffer.dall@linaro.org>

Moving the call to vcpu_load() in kvm_arch_vcpu_ioctl_run() to after
we've called kvm_vcpu_first_run_init() simplifies some of the vgic and
there is also no need to do vcpu_load() for things such as handling the
immediate_exit flag.

Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
---
 virt/kvm/arm/arch_timer.c     |  4 ----
 virt/kvm/arm/arm.c            | 12 +++++-------
 virt/kvm/arm/vgic/vgic-init.c | 11 -----------
 3 files changed, 5 insertions(+), 22 deletions(-)

diff --git a/virt/kvm/arm/arch_timer.c b/virt/kvm/arm/arch_timer.c
index 4151250ce8da..801fecfee299 100644
--- a/virt/kvm/arm/arch_timer.c
+++ b/virt/kvm/arm/arch_timer.c
@@ -839,11 +839,7 @@ int kvm_timer_enable(struct kvm_vcpu *vcpu)
 		return ret;
 
 no_vgic:
-	preempt_disable();
 	timer->enabled = 1;
-	kvm_timer_vcpu_load_vgic(vcpu);
-	preempt_enable();
-
 	return 0;
 }
 
diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
index a760ef1803be..991f1aa70fb9 100644
--- a/virt/kvm/arm/arm.c
+++ b/virt/kvm/arm/arm.c
@@ -622,8 +622,6 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (unlikely(!kvm_vcpu_initialized(vcpu)))
 		return -ENOEXEC;
 
-	vcpu_load(vcpu);
-
 	ret = kvm_vcpu_first_run_init(vcpu);
 	if (ret)
 		goto out;
@@ -631,13 +629,13 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu, struct kvm_run *run)
 	if (run->exit_reason == KVM_EXIT_MMIO) {
 		ret = kvm_handle_mmio_return(vcpu, vcpu->run);
 		if (ret)
-			goto out;
+			return ret;
 	}
 
-	if (run->immediate_exit) {
-		ret = -EINTR;
-		goto out;
-	}
+	if (run->immediate_exit)
+		return -EINTR;
+
+	vcpu_load(vcpu);
 
 	if (vcpu->sigset_active)
 		sigprocmask(SIG_SETMASK, &vcpu->sigset, &sigsaved);
diff --git a/virt/kvm/arm/vgic/vgic-init.c b/virt/kvm/arm/vgic/vgic-init.c
index 62310122ee78..a0688ef52ad7 100644
--- a/virt/kvm/arm/vgic/vgic-init.c
+++ b/virt/kvm/arm/vgic/vgic-init.c
@@ -300,17 +300,6 @@ int vgic_init(struct kvm *kvm)
 
 	dist->initialized = true;
 
-	/*
-	 * If we're initializing GICv2 on-demand when first running the VCPU
-	 * then we need to load the VGIC state onto the CPU.  We can detect
-	 * this easily by checking if we are in between vcpu_load and vcpu_put
-	 * when we just initialized the VGIC.
-	 */
-	preempt_disable();
-	vcpu = kvm_arm_get_running_vcpu();
-	if (vcpu)
-		kvm_vgic_load(vcpu);
-	preempt_enable();
 out:
 	return ret;
 }
-- 
2.14.2