Skip to contentSkip to conversation

Target CPU=Host implementation for KVM ARM/ARM64

PATCH Superseded Anup Patel 3 participants 7 messages Sep 11, 2013

Subsystems (3)

v1 of 5 Superseded Latest is v5 →

lore
Share & tools
lore

Anup Patel

author
#0 lore raw

cover Target CPU=Host implementation for KVM ARM/ARM64

It will be very useful for user space if it has a method of querying VCPU target type suitable to underlying Host. We can use such querying mechanism and implement machine models in user space where VCPU target type is always same-as/suitable-to underlying Host (i.e. Target CPU=Host).

This patch series implements KVM_ARM_SUITABLE_TARGET ioclt for querying VCPU target type matching underlying host. Using this new ioctl we can implement VCPU target CPU=Host in user space (i.e. QEMU/KVMTOOL).

Also, it is not mandatory to call KVM_ARM_SUITABLE_TARGET ioctl and the old method of using only KVM_ARM_VCPU_INIT ioctl works fine.

Anup Patel (2):
ARM/ARM64: KVM: Implement KVM_ARM_SUITABLE_TARGET ioctl
KVM: Add documentation for KVM_ARM_SUITABLE_TARGET ioctl

 Documentation/virtual/kvm/api.txt |   25 ++++++++++++++++++++-----
 arch/arm/kvm/arm.c                |    2 ++
 include/uapi/linux/kvm.h          |    1 +
 3 files changed, 23 insertions(+), 5 deletions(-)
Signature · 1 line
1.7.9.5
author
#1 lore raw

PATCH 1/2 ARM/ARM64: KVM: Implement KVM_ARM_SUITABLE_TARGET ioctl

For implementing CPU=host, we need a mechanism for querying VCPU target type suitable to underlying Host.

This patch implements KVM_ARM_SUITABLE_TARGET ioctl which return output of kvm_arm_target() to user space for KVM ARM and KVM ARM64.

---
 arch/arm/kvm/arm.c       |    2 ++
 include/uapi/linux/kvm.h |    1 +
 2 files changed, 3 insertions(+)
arch/arm/kvm/arm.c +2 -0
index 9c697db..4bdc2e1 100644
@@ -715,6 +715,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		return kvm_vcpu_set_target(vcpu, &init);
 
 	}
+	case KVM_ARM_SUITABLE_TARGET:
+		return kvm_target_cpu();
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
include/uapi/linux/kvm.h +1 -0
index 99c2533..8285daa 100644
@@ -1012,6 +1012,7 @@ struct kvm_s390_ucas_mapping {
 /* VM is being stopped by host */
 #define KVM_KVMCLOCK_CTRL	  _IO(KVMIO,   0xad)
 #define KVM_ARM_VCPU_INIT	  _IOW(KVMIO,  0xae, struct kvm_vcpu_init)
+#define KVM_ARM_SUITABLE_TARGET	  _IO(KVMIO,   0xaf)
 #define KVM_GET_REG_LIST	  _IOWR(KVMIO, 0xb0, struct kvm_reg_list)
 
 #define KVM_DEV_ASSIGN_ENABLE_IOMMU	(1 << 0)
Signature · 1 line
1.7.9.5
#2 lore raw

PATCH 1/2 ARM/ARM64: KVM: Implement KVM_ARM_SUITABLE_TARGET ioctl

Quoted 19 lines · Anup Patel

For implementing CPU=host, we need a mechanism for querying VCPU target type suitable to underlying Host.

This patch implements KVM_ARM_SUITABLE_TARGET ioctl which return output of kvm_arm_target() to user space for KVM ARM and KVM ARM64.

---
 arch/arm/kvm/arm.c       |    2 ++
 include/uapi/linux/kvm.h |    1 +
 2 files changed, 3 insertions(+)
arch/arm/kvm/arm.c +0 -0
index 9c697db..4bdc2e1 100644

arch/arm/kvm/arm.c long kvm_arch_vcpu_ioctl(struct file *filp, · line 719

Quoted 5 lines
arch/arm/kvm/arm.c +2 -0
@@ -715,6 +715,8 @@ long kvm_arch_vcpu_ioctl(struct file *filp,
 		return kvm_vcpu_set_target(vcpu, &init);
 
 	}
+	case KVM_ARM_SUITABLE_TARGET:
+		return kvm_target_cpu();

as I commented on the other patch I think we should try to think of a better name and this shouldn't be a VCPU ioctl.

Quoted 18 lines
arch/arm/kvm/arm.c +0 -0
 	case KVM_SET_ONE_REG:
 	case KVM_GET_ONE_REG: {
 		struct kvm_one_reg reg;
include/uapi/linux/kvm.h +1 -0
index 99c2533..8285daa 100644
@@ -1012,6 +1012,7 @@ struct kvm_s390_ucas_mapping {
 /* VM is being stopped by host */
 #define KVM_KVMCLOCK_CTRL	  _IO(KVMIO,   0xad)
 #define KVM_ARM_VCPU_INIT	  _IOW(KVMIO,  0xae, struct kvm_vcpu_init)
+#define KVM_ARM_SUITABLE_TARGET	  _IO(KVMIO,   0xaf)
 #define KVM_GET_REG_LIST	  _IOWR(KVMIO, 0xb0, struct kvm_reg_list)
 
 #define KVM_DEV_ASSIGN_ENABLE_IOMMU	(1 << 0)
Signature · 1 line
1.7.9.5
author
#3 lore raw

PATCH 2/2 KVM: Add documentation for KVM_ARM_SUITABLE_TARGET ioctl

In reply to Anup Patel

To implement CPU=Host we have added KVM_ARM_SUITABLE_TARGET ioctl which provides a CPU target type to user space for creating VCPU matching underlying Host.

This patch adds info related to this new KVM_ARM_SUITABLE_TARGET ioctl in the KVM API documentation.

---
 Documentation/virtual/kvm/api.txt |   25 ++++++++++++++++++++-----
 1 file changed, 20 insertions(+), 5 deletions(-)
Documentation/virtual/kvm/api.txt +20 -5
index ef925ea..1ae9721 100644
@@ -2283,7 +2283,7 @@ current state.  "addr" is ignored.
 Capability: basic
 Architectures: arm, arm64
 Type: vcpu ioctl
-Parameters: struct struct kvm_vcpu_init (in)
+Parameters: struct kvm_vcpu_init (in)
 Returns: 0 on success; -1 on error
 Errors:
  ?EINVAL: ???the target is unknown, or the combination of features is invalid.
@@ -2303,8 +2303,24 @@ Possible features:
 	- KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode.
 	  Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only).
 
+4.83 KVM_ARM_SUITABLE_TARGET
 
-4.83 KVM_GET_REG_LIST
+Capability: basic
+Architectures: arm, arm64
+Type: vcpu ioctl
+Parameters: None
+Returns: 0 on success; -1 on error
+Errors:
+ ?EINVAL: ???no suitable target available for the host
+
+This queries KVM for suitable CPU target type which can be emulated by 
+KVM on underlying host. This is not a mandatory API and could be used
+to create VCPUs matching underlying host.
+
+The ioctl returns a target type which can be directly passed-back to 
+the KVM_ARM_VCPU_INIT ioctl.
+
+4.84 KVM_GET_REG_LIST
 
 Capability: basic
 Architectures: arm, arm64
@@ -2323,8 +2339,7 @@ struct kvm_reg_list {
 This ioctl returns the guest registers that are supported for the
 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
 
-
-4.84 KVM_ARM_SET_DEVICE_ADDR
+4.85 KVM_ARM_SET_DEVICE_ADDR
 
 Capability: KVM_CAP_ARM_SET_DEVICE_ADDR
 Architectures: arm, arm64
@@ -2362,7 +2377,7 @@ must be called after calling KVM_CREATE_IRQCHIP, but before calling
 KVM_RUN on any of the VCPUs.  Calling this ioctl twice for any of the
 base addresses will return -EEXIST.
 
-4.85 KVM_PPC_RTAS_DEFINE_TOKEN
+4.86 KVM_PPC_RTAS_DEFINE_TOKEN
 
 Capability: KVM_CAP_PPC_RTAS
 Architectures: ppc
Signature · 1 line
1.7.9.5
#4 lore raw

PATCH 2/2 KVM: Add documentation for KVM_ARM_SUITABLE_TARGET ioctl

Quoted 27 lines · Anup Patel

To implement CPU=Host we have added KVM_ARM_SUITABLE_TARGET ioctl which provides a CPU target type to user space for creating VCPU matching underlying Host.

This patch adds info related to this new KVM_ARM_SUITABLE_TARGET ioctl in the KVM API documentation.

---
 Documentation/virtual/kvm/api.txt |   25 ++++++++++++++++++++-----
 1 file changed, 20 insertions(+), 5 deletions(-)
Documentation/virtual/kvm/api.txt +1 -1
index ef925ea..1ae9721 100644
@@ -2283,7 +2283,7 @@ current state.  "addr" is ignored.
 Capability: basic
 Architectures: arm, arm64
 Type: vcpu ioctl
-Parameters: struct struct kvm_vcpu_init (in)
+Parameters: struct kvm_vcpu_init (in)
 Returns: 0 on success; -1 on error
 Errors:
  ?EINVAL: ???the target is unknown, or the combination of features is invalid.

Documentation/virtual/kvm/api.txt Possible features: · line 2304

Quoted 2 lines
Documentation/virtual/kvm/api.txt +0 -0
@@ -2303,8 +2303,24 @@ Possible features:
 	- KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode.
 	  Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only).

unrelated change, can you put in separate patch?

Documentation/virtual/kvm/api.txt Possible features: · line 2306

Quoted 2 lines
Documentation/virtual/kvm/api.txt +1 -0
+4.83 KVM_ARM_SUITABLE_TARGET

hmm, I'm a little uneasy about the use of the word 'suitable'. KVM_ARM_PREFERRED_TARGET ?

Documentation/virtual/kvm/api.txt Possible features: · line 2310

Quoted 5 lines
Documentation/virtual/kvm/api.txt +3 -1
-4.83 KVM_GET_REG_LIST
+Capability: basic
+Architectures: arm, arm64
+Type: vcpu ioctl

this should be a system ioctl (or at least a VM ioctl) so we can query the preferred vcpu without creating one.

Documentation/virtual/kvm/api.txt Possible features: · line 2312

Quoted 2 lines
Documentation/virtual/kvm/api.txt +2 -0
+Parameters: None
+Returns: 0 on success; -1 on error

no parameters and return 0 on success?

I would suggest returning struct kvm_vcpu_init (out) instead. It makes for a more flexible API in case we want to say something about the preferred/supported vcpu features as well in the future.

Documentation/virtual/kvm/api.txt Possible features: · line 2314

Quoted 2 lines
Documentation/virtual/kvm/api.txt +2 -0
+Errors:
+ ?EINVAL: ???no suitable target available for the host

EINVAL is not really a proper error code for this IMHO, because it indicates the user supplied something incorrect. How about ENODEV?

Documentation/virtual/kvm/api.txt Possible features: · line 2318

Quoted 4 lines
Documentation/virtual/kvm/api.txt +4 -0
+
+This queries KVM for suitable CPU target type which can be emulated by 
+KVM on underlying host. This is not a mandatory API and could be used
+to create VCPUs matching underlying host.

this last sentece caused me to trip. Isn't the second part of the sentence (after the 'and') what you are trying to say in the first sentece, and what you want to say here is that it is not required to call this ioctl before calling KVM_ARM_VCPU_INIT? I'm not sure this information is required at all, but it doesn't hurt I guess.

Documentation/virtual/kvm/api.txt Possible features: · line 2318

Quoted 2 lines
Documentation/virtual/kvm/api.txt +2 -0
+
+The ioctl returns a target type which can be directly passed-back to 

s/passed-back/passed back/
or consider revising it to "... which can be used for the target type in the KVM_ARM_VCPU_INIT ioctl."

Quoted 28 lines
Documentation/virtual/kvm/api.txt +5 -3
+the KVM_ARM_VCPU_INIT ioctl.
+
+4.84 KVM_GET_REG_LIST
 
 Capability: basic
 Architectures: arm, arm64
@@ -2323,8 +2339,7 @@ struct kvm_reg_list {
 This ioctl returns the guest registers that are supported for the
 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
 
-
-4.84 KVM_ARM_SET_DEVICE_ADDR
+4.85 KVM_ARM_SET_DEVICE_ADDR
 
 Capability: KVM_CAP_ARM_SET_DEVICE_ADDR
 Architectures: arm, arm64
@@ -2362,7 +2377,7 @@ must be called after calling KVM_CREATE_IRQCHIP, but before calling
 KVM_RUN on any of the VCPUs.  Calling this ioctl twice for any of the
 base addresses will return -EEXIST.
 
-4.85 KVM_PPC_RTAS_DEFINE_TOKEN
+4.86 KVM_PPC_RTAS_DEFINE_TOKEN
 
 Capability: KVM_CAP_PPC_RTAS
 Architectures: ppc
Signature · 1 line
1.7.9.5
#5 lore raw

PATCH 2/2 KVM: Add documentation for KVM_ARM_SUITABLE_TARGET ioctl

Quoted 24 lines · Christoffer Dall
Quoted 23 lines · Anup Patel

To implement CPU=Host we have added KVM_ARM_SUITABLE_TARGET ioctl which provides a CPU target type to user space for creating VCPU matching underlying Host.

This patch adds info related to this new KVM_ARM_SUITABLE_TARGET ioctl in the KVM API documentation.

---
 Documentation/virtual/kvm/api.txt |   25 ++++++++++++++++++++-----
 1 file changed, 20 insertions(+), 5 deletions(-)
Documentation/virtual/kvm/api.txt +1 -1
index ef925ea..1ae9721 100644
@@ -2283,7 +2283,7 @@ current state.  "addr" is ignored.
 Capability: basic
 Architectures: arm, arm64
 Type: vcpu ioctl
-Parameters: struct struct kvm_vcpu_init (in)
+Parameters: struct kvm_vcpu_init (in)
Quoted 8 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2304

Quoted 6 lines
Documentation/virtual/kvm/api.txt +0 -0
 Returns: 0 on success; -1 on error
 Errors:
   EINVAL:    the target is unknown, or the combination of features is invalid.
@@ -2303,8 +2303,24 @@ Possible features:
      - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode.
        Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only).

unrelated change, can you put in separate patch?

Ok

Quoted 6 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2306

Quoted 3 lines
Documentation/virtual/kvm/api.txt +1 -0
+4.83 KVM_ARM_SUITABLE_TARGET

hmm, I'm a little uneasy about the use of the word 'suitable'. KVM_ARM_PREFERRED_TARGET ?

Sure, I will rename it to KVM_ARM_PREFERRED_TARGET.

Quoted 9 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2310

Quoted 6 lines
Documentation/virtual/kvm/api.txt +3 -1
-4.83 KVM_GET_REG_LIST
+Capability: basic
+Architectures: arm, arm64
+Type: vcpu ioctl

this should be a system ioctl (or at least a VM ioctl) so we can query the preferred vcpu without creating one.

I was really confused on this one because the ioctl is all about VCPU target type but I am fine calling it system (or VM) ioctl.

Quoted 5 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2312

Quoted 3 lines
Documentation/virtual/kvm/api.txt +2 -0
+Parameters: None
+Returns: 0 on success; -1 on error

no parameters and return 0 on success?

Typo from my-side.
It returns positive value upon success.

Quoted 4 lines

I would suggest returning struct kvm_vcpu_init (out) instead. It makes for a more flexible API in case we want to say something about the preferred/supported vcpu features as well in the future.

I started-off with an implementation which returned struct kvm_vcpu_init but there will be features such POWER_OFF which be only decided by user space.

For e.g. SMP Guest might start VCPU0 in POWER_ON whereas other VCPUs in POWER_OFF. We cannot suggest anything about POWER state of VCPU to user space from KVM because we really dont know what machine model is emulated by user space.

IMHO, the only thing that seems worth returning is the VCPU target type hence this implementation.

Quoted 6 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2312

Quoted 3 lines
Documentation/virtual/kvm/api.txt +2 -0
+Errors:
+  EINVAL:    no suitable target available for the host

EINVAL is not really a proper error code for this IMHO, because it indicates the user supplied something incorrect. How about ENODEV?

Ok, I can change this.

I kept EINVAL because I did not want to make changes in kvm_target_cpu().

Quoted 11 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2318

Quoted 5 lines
Documentation/virtual/kvm/api.txt +4 -0
+
+This queries KVM for suitable CPU target type which can be emulated by
+KVM on underlying host. This is not a mandatory API and could be used
+to create VCPUs matching underlying host.

this last sentece caused me to trip. Isn't the second part of the sentence (after the 'and') what you are trying to say in the first sentece, and what you want to say here is that it is not required to call this ioctl before calling KVM_ARM_VCPU_INIT? I'm not sure this information is required at all, but it doesn't hurt I guess.

Ok, let me try to re-phrase this.

Quoted 7 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2318

Quoted 3 lines
Documentation/virtual/kvm/api.txt +2 -0
+
+The ioctl returns a target type which can be directly passed-back to

s/passed-back/passed back/
or consider revising it to "... which can be used for the target type in the KVM_ARM_VCPU_INIT ioctl."

Ok

Quoted 33 lines
Quoted 27 lines
Documentation/virtual/kvm/api.txt +5 -4
+the KVM_ARM_VCPU_INIT ioctl.
+
+4.84 KVM_GET_REG_LIST
 
 Capability: basic
 Architectures: arm, arm64
@@ -2323,8 +2339,7 @@ struct kvm_reg_list {
 This ioctl returns the guest registers that are supported for the
 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
 
-
-4.84 KVM_ARM_SET_DEVICE_ADDR
+4.85 KVM_ARM_SET_DEVICE_ADDR
 
 Capability: KVM_CAP_ARM_SET_DEVICE_ADDR
 Architectures: arm, arm64
@@ -2362,7 +2377,7 @@ must be called after calling KVM_CREATE_IRQCHIP, but before calling
 KVM_RUN on any of the VCPUs.  Calling this ioctl twice for any of the
 base addresses will return -EEXIST.
 
-4.85 KVM_PPC_RTAS_DEFINE_TOKEN
+4.86 KVM_PPC_RTAS_DEFINE_TOKEN
 
 Capability: KVM_CAP_PPC_RTAS
 Architectures: ppc
--
Quoted 2 lines

1.7.9.5

_______________________________________________
kvmarm mailing list
kvmarm at lists.cs.columbia.edu
https://lists.cs.columbia.edu/cucslists/listinfo/kvmarm

#6 lore raw

PATCH 2/2 KVM: Add documentation for KVM_ARM_SUITABLE_TARGET ioctl

Quoted 56 lines · Anup Patel
Quoted 24 lines · Christoffer Dall
Quoted 23 lines · Anup Patel

To implement CPU=Host we have added KVM_ARM_SUITABLE_TARGET ioctl which provides a CPU target type to user space for creating VCPU matching underlying Host.

This patch adds info related to this new KVM_ARM_SUITABLE_TARGET ioctl in the KVM API documentation.

---
 Documentation/virtual/kvm/api.txt |   25 ++++++++++++++++++++-----
 1 file changed, 20 insertions(+), 5 deletions(-)
Documentation/virtual/kvm/api.txt +1 -1
index ef925ea..1ae9721 100644
@@ -2283,7 +2283,7 @@ current state.  "addr" is ignored.
 Capability: basic
 Architectures: arm, arm64
 Type: vcpu ioctl
-Parameters: struct struct kvm_vcpu_init (in)
+Parameters: struct kvm_vcpu_init (in)
Quoted 8 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2304

Quoted 6 lines
Documentation/virtual/kvm/api.txt +0 -0
 Returns: 0 on success; -1 on error
 Errors:
   EINVAL:    the target is unknown, or the combination of features is invalid.
@@ -2303,8 +2303,24 @@ Possible features:
      - KVM_ARM_VCPU_EL1_32BIT: Starts the CPU in a 32bit mode.
        Depends on KVM_CAP_ARM_EL1_32BIT (arm64 only).

unrelated change, can you put in separate patch?

Ok

Quoted 6 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2306

Quoted 3 lines
Documentation/virtual/kvm/api.txt +1 -0
+4.83 KVM_ARM_SUITABLE_TARGET

hmm, I'm a little uneasy about the use of the word 'suitable'. KVM_ARM_PREFERRED_TARGET ?

Sure, I will rename it to KVM_ARM_PREFERRED_TARGET.

Quoted 9 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2310

Quoted 6 lines
Documentation/virtual/kvm/api.txt +3 -1
-4.83 KVM_GET_REG_LIST
+Capability: basic
+Architectures: arm, arm64
+Type: vcpu ioctl

this should be a system ioctl (or at least a VM ioctl) so we can query the preferred vcpu without creating one.

Quoted 3 lines

I was really confused on this one because the ioctl is all about VCPU target type but I am fine calling it system (or VM) ioctl.

the 'Type' goes to which type of file descriptor the ioctl is called on. After I wrote the comment I realized you don't technically need the type of the CPU before creating a vcpu fd, but I think it's nicer this way, so user space can just tell at the earliest state what this particular KVM instance prefers.

Quoted 18 lines
Quoted 5 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2312

Quoted 3 lines
Documentation/virtual/kvm/api.txt +2 -0
+Parameters: None
+Returns: 0 on success; -1 on error

no parameters and return 0 on success?

Typo from my-side.
It returns positive value upon success.

Quoted 4 lines

I would suggest returning struct kvm_vcpu_init (out) instead. It makes for a more flexible API in case we want to say something about the preferred/supported vcpu features as well in the future.

I started-off with an implementation which returned struct kvm_vcpu_init but there will be features such POWER_OFF which be only decided by user space.

Quoted 8 lines

For e.g. SMP Guest might start VCPU0 in POWER_ON whereas other VCPUs in POWER_OFF. We cannot suggest anything about POWER state of VCPU to user space from KVM because we really dont know what machine model is emulated by user space.

IMHO, the only thing that seems worth returning is the VCPU target type hence this implementation.

I agree as things are looking right now, but I wonder if we would later add cpu features like an emulated SMC environment and we want to be able to return that information without having to come up with yet another ioctl. Are there any downsides in returning the kvm_vcpu_init struct and just specify here in the docs that the features field does not return anything useful (yet)?

Quoted 4 lines
Quoted 4 lines
Quoted 3 lines
Documentation/virtual/kvm/api.txt +2 -0
+Errors:
+  EINVAL:    no suitable target available for the host
Quoted 8 lines
Quoted 2 lines

EINVAL is not really a proper error code for this IMHO, because it indicates the user supplied something incorrect. How about ENODEV?

Ok, I can change this.

I kept EINVAL because I did not want to make changes in kvm_target_cpu().

ah, ok, yeah, the EINVAL tends to be too overloaded inside the kernel and I guess I fell into that trap as well. If it doesn't make the code too ugly, I think another error would be bette for a user space API.

Should be fixable with something like

return (ret == -EINVAL) ? -ENODEV : ret;

Quoted 57 lines
Quoted 11 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2318

Quoted 5 lines
Documentation/virtual/kvm/api.txt +4 -0
+
+This queries KVM for suitable CPU target type which can be emulated by
+KVM on underlying host. This is not a mandatory API and could be used
+to create VCPUs matching underlying host.

this last sentece caused me to trip. Isn't the second part of the sentence (after the 'and') what you are trying to say in the first sentece, and what you want to say here is that it is not required to call this ioctl before calling KVM_ARM_VCPU_INIT? I'm not sure this information is required at all, but it doesn't hurt I guess.

Ok, let me try to re-phrase this.

Quoted 7 lines

Documentation/virtual/kvm/api.txt Possible features: · line 2318

Quoted 3 lines
Documentation/virtual/kvm/api.txt +2 -0
+
+The ioctl returns a target type which can be directly passed-back to

s/passed-back/passed back/
or consider revising it to "... which can be used for the target type in the KVM_ARM_VCPU_INIT ioctl."

Ok

Quoted 33 lines
Quoted 27 lines
Documentation/virtual/kvm/api.txt +5 -4
+the KVM_ARM_VCPU_INIT ioctl.
+
+4.84 KVM_GET_REG_LIST
 
 Capability: basic
 Architectures: arm, arm64
@@ -2323,8 +2339,7 @@ struct kvm_reg_list {
 This ioctl returns the guest registers that are supported for the
 KVM_GET_ONE_REG/KVM_SET_ONE_REG calls.
 
-
-4.84 KVM_ARM_SET_DEVICE_ADDR
+4.85 KVM_ARM_SET_DEVICE_ADDR
 
 Capability: KVM_CAP_ARM_SET_DEVICE_ADDR
 Architectures: arm, arm64
@@ -2362,7 +2377,7 @@ must be called after calling KVM_CREATE_IRQCHIP, but before calling
 KVM_RUN on any of the VCPUs.  Calling this ioctl twice for any of the
 base addresses will return -EEXIST.
 
-4.85 KVM_PPC_RTAS_DEFINE_TOKEN
+4.86 KVM_PPC_RTAS_DEFINE_TOKEN
 
 Capability: KVM_CAP_PPC_RTAS
 Architectures: ppc
--
Quoted 2 lines

1.7.9.5

_______________________________________________
kvmarm mailing list
kvmarm at lists.cs.columbia.edu
https://lists.cs.columbia.edu/cucslists/listinfo/kvmarm