Skip to contentSkip to conversation

kvm: Remove KVM_CREATE_IRQCHIP support assumption

PATCH Superseded Andrew Jones 2 participants 4 messages Jul 22, 2023 – Jul 25, 2023 KVM

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 kvm: Remove KVM_CREATE_IRQCHIP support assumption

Since Linux commit 00f918f61c56 ("RISC-V: KVM: Skeletal in-kernel AIA irqchip support") checking KVM_CAP_IRQCHIP returns non-zero when the RISC-V platform has AIA. The cap indicates KVM supports at least one of the following ioctls:

KVM_CREATE_IRQCHIP
KVM_IRQ_LINE
KVM_GET_IRQCHIP
KVM_SET_IRQCHIP
KVM_GET_LAPIC
KVM_SET_LAPIC

but the cap doesn't imply that KVM must support any of those ioctls in particular. However, QEMU was assuming the KVM_CREATE_IRQCHIP ioctl was supported. Stop making that assumption by introducing a KVM parameter that each architecture which supports KVM_CREATE_IRQCHIP sets. Adding parameters isn't awesome, but given how the KVM_CAP_IRQCHIP isn't very helpful on its own, we don't have a lot of options.

---

While this fixes booting guests on riscv KVM with AIA it's unlikely to get merged before the QEMU support for KVM AIA[1] lands, which would also fix the issue. I think this patch is still worth considering though since QEMU's assumption is wrong.

[1] https://lore.kernel.org/all/[email protected]/

 accel/kvm/kvm-all.c    | 5 ++++-
 include/sysemu/kvm.h   | 1 +
 target/arm/kvm.c       | 3 +++
 target/i386/kvm/kvm.c  | 2 ++
 target/s390x/kvm/kvm.c | 3 +++
 5 files changed, 13 insertions(+), 1 deletion(-)
accel/kvm/kvm-all.c +4 -1
index 373d876c0580..0f5ff8630502 100644
@@ -86,6 +86,7 @@ struct KVMParkedVcpu {
 };
 
 KVMState *kvm_state;
+bool kvm_has_create_irqchip;
 bool kvm_kernel_irqchip;
 bool kvm_split_irqchip;
 bool kvm_async_interrupts_allowed;
@@ -2377,8 +2378,10 @@ static void kvm_irqchip_create(KVMState *s)
         if (s->kernel_irqchip_split == ON_OFF_AUTO_ON) {
             error_report("Split IRQ chip mode not supported.");
             exit(1);
-        } else {
+        } else if (kvm_has_create_irqchip) {
             ret = kvm_vm_ioctl(s, KVM_CREATE_IRQCHIP);
+        } else {
+            return;
         }
     }
     if (ret < 0) {
include/sysemu/kvm.h +1 -0
index 115f0cca79d1..84b1bb3dc91e 100644
@@ -32,6 +32,7 @@
 #ifdef CONFIG_KVM_IS_POSSIBLE
 
 extern bool kvm_allowed;
+extern bool kvm_has_create_irqchip;
 extern bool kvm_kernel_irqchip;
 extern bool kvm_split_irqchip;
 extern bool kvm_async_interrupts_allowed;
target/arm/kvm.c +3 -0
index b4c7654f4980..2fa87b495d68 100644
@@ -250,6 +250,9 @@ int kvm_arm_get_max_vm_ipa_size(MachineState *ms, bool *fixed_ipa)
 int kvm_arch_init(MachineState *ms, KVMState *s)
 {
     int ret = 0;
+
+    kvm_has_create_irqchip = kvm_check_extension(s, KVM_CAP_IRQCHIP);
+
     /* For ARM interrupt delivery is always asynchronous,
      * whether we are using an in-kernel VGIC or not.
      */
target/i386/kvm/kvm.c +2 -0
index ebfaf3d24c79..6363e67f092d 100644
@@ -2771,6 +2771,8 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
         }
     }
 
+    kvm_has_create_irqchip = kvm_check_extension(s, KVM_CAP_IRQCHIP);
+
     return 0;
 }
 
target/s390x/kvm/kvm.c +3 -0
index a9e5880349d9..c053304adf94 100644
@@ -391,6 +391,9 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
     }
 
     kvm_set_max_memslot_size(KVM_SLOT_MAX_BYTES);
+
+    kvm_has_create_irqchip = kvm_check_extension(s, KVM_CAP_S390_IRQCHIP);
+
     return 0;
 }
 
Signature · 1 line
2.41.0
#1 lore raw
Quoted 52 lines · Andrew Jones

Since Linux commit 00f918f61c56 ("RISC-V: KVM: Skeletal in-kernel AIA irqchip support") checking KVM_CAP_IRQCHIP returns non-zero when the RISC-V platform has AIA. The cap indicates KVM supports at least one of the following ioctls:

KVM_CREATE_IRQCHIP
KVM_IRQ_LINE
KVM_GET_IRQCHIP
KVM_SET_IRQCHIP
KVM_GET_LAPIC
KVM_SET_LAPIC

but the cap doesn't imply that KVM must support any of those ioctls in particular. However, QEMU was assuming the KVM_CREATE_IRQCHIP ioctl was supported. Stop making that assumption by introducing a KVM parameter that each architecture which supports KVM_CREATE_IRQCHIP sets. Adding parameters isn't awesome, but given how the KVM_CAP_IRQCHIP isn't very helpful on its own, we don't have a lot of options.

---

While this fixes booting guests on riscv KVM with AIA it's unlikely to get merged before the QEMU support for KVM AIA[1] lands, which would also fix the issue. I think this patch is still worth considering though since QEMU's assumption is wrong.

[1] https://lore.kernel.org/all/[email protected]/

  accel/kvm/kvm-all.c    | 5 ++++-
  include/sysemu/kvm.h   | 1 +
  target/arm/kvm.c       | 3 +++
  target/i386/kvm/kvm.c  | 2 ++
  target/s390x/kvm/kvm.c | 3 +++
  5 files changed, 13 insertions(+), 1 deletion(-)
accel/kvm/kvm-all.c +1 -0
index 373d876c0580..0f5ff8630502 100644
@@ -86,6 +86,7 @@ struct KVMParkedVcpu {
  };
 
  KVMState *kvm_state;
+bool kvm_has_create_irqchip;
  bool kvm_kernel_irqchip;
  bool kvm_split_irqchip;
  bool kvm_async_interrupts_allowed;
@@ -2377,8 +2378,10 @@ static void kvm_irqchip_create(KVMState *s)
          if (s->kernel_irqchip_split == ON_OFF_AUTO_ON) {

accel/kvm/kvm-all.c static void kvm_irqchip_create(KVMState *s) · line 2387

Quoted 10 lines
accel/kvm/kvm-all.c +3 -1
              error_report("Split IRQ chip mode not supported.");
              exit(1);
-        } else {
+        } else if (kvm_has_create_irqchip) {
              ret = kvm_vm_ioctl(s, KVM_CREATE_IRQCHIP);
+        } else {
+            return;
          }
      }
      if (ret < 0) {

I think I'd do this differntly... at the beginning of the function, there is a check for kvm_check_extension(s, KVM_CAP_IRQCHIP) etc. ... I think you could now replace that check with a simple

	if (!kvm_has_create_irqchip) {
		return;
	}

The "kvm_vm_enable_cap(s, KVM_CAP_S390_IRQCHIP, 0)" of course has to be moved to the target/s390x/kvm/kvm.c file, too.

Thomas

Quoted 53 lines
include/sysemu/kvm.h +1 -0
index 115f0cca79d1..84b1bb3dc91e 100644
@@ -32,6 +32,7 @@
  #ifdef CONFIG_KVM_IS_POSSIBLE
 
  extern bool kvm_allowed;
+extern bool kvm_has_create_irqchip;
  extern bool kvm_kernel_irqchip;
  extern bool kvm_split_irqchip;
  extern bool kvm_async_interrupts_allowed;
target/arm/kvm.c +3 -0
index b4c7654f4980..2fa87b495d68 100644
@@ -250,6 +250,9 @@ int kvm_arm_get_max_vm_ipa_size(MachineState *ms, bool *fixed_ipa)
  int kvm_arch_init(MachineState *ms, KVMState *s)
  {
      int ret = 0;
+
+    kvm_has_create_irqchip = kvm_check_extension(s, KVM_CAP_IRQCHIP);
+
      /* For ARM interrupt delivery is always asynchronous,
       * whether we are using an in-kernel VGIC or not.
       */
target/i386/kvm/kvm.c +2 -0
index ebfaf3d24c79..6363e67f092d 100644
@@ -2771,6 +2771,8 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
          }
      }
 
+    kvm_has_create_irqchip = kvm_check_extension(s, KVM_CAP_IRQCHIP);
+
      return 0;
  }
 
target/s390x/kvm/kvm.c +3 -0
index a9e5880349d9..c053304adf94 100644
@@ -391,6 +391,9 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
      }
 
      kvm_set_max_memslot_size(KVM_SLOT_MAX_BYTES);
+
+    kvm_has_create_irqchip = kvm_check_extension(s, KVM_CAP_S390_IRQCHIP);
+
      return 0;
  }
 
#2 lore raw
Quoted 67 lines · Thomas Huth
Quoted 51 lines · Andrew Jones

Since Linux commit 00f918f61c56 ("RISC-V: KVM: Skeletal in-kernel AIA irqchip support") checking KVM_CAP_IRQCHIP returns non-zero when the RISC-V platform has AIA. The cap indicates KVM supports at least one of the following ioctls:

KVM_CREATE_IRQCHIP
KVM_IRQ_LINE
KVM_GET_IRQCHIP
KVM_SET_IRQCHIP
KVM_GET_LAPIC
KVM_SET_LAPIC

but the cap doesn't imply that KVM must support any of those ioctls in particular. However, QEMU was assuming the KVM_CREATE_IRQCHIP ioctl was supported. Stop making that assumption by introducing a KVM parameter that each architecture which supports KVM_CREATE_IRQCHIP sets. Adding parameters isn't awesome, but given how the KVM_CAP_IRQCHIP isn't very helpful on its own, we don't have a lot of options.

---

While this fixes booting guests on riscv KVM with AIA it's unlikely to get merged before the QEMU support for KVM AIA[1] lands, which would also fix the issue. I think this patch is still worth considering though since QEMU's assumption is wrong.

[1] https://lore.kernel.org/all/[email protected]/

  accel/kvm/kvm-all.c    | 5 ++++-
  include/sysemu/kvm.h   | 1 +
  target/arm/kvm.c       | 3 +++
  target/i386/kvm/kvm.c  | 2 ++
  target/s390x/kvm/kvm.c | 3 +++
  5 files changed, 13 insertions(+), 1 deletion(-)
accel/kvm/kvm-all.c +1 -0
index 373d876c0580..0f5ff8630502 100644
@@ -86,6 +86,7 @@ struct KVMParkedVcpu {
  };
  KVMState *kvm_state;
+bool kvm_has_create_irqchip;
  bool kvm_kernel_irqchip;
  bool kvm_split_irqchip;
  bool kvm_async_interrupts_allowed;
@@ -2377,8 +2378,10 @@ static void kvm_irqchip_create(KVMState *s)
          if (s->kernel_irqchip_split == ON_OFF_AUTO_ON) {

accel/kvm/kvm-all.c static void kvm_irqchip_create(KVMState *s) · line 2387

Quoted 10 lines
accel/kvm/kvm-all.c +3 -1
              error_report("Split IRQ chip mode not supported.");
              exit(1);
-        } else {
+        } else if (kvm_has_create_irqchip) {
              ret = kvm_vm_ioctl(s, KVM_CREATE_IRQCHIP);
+        } else {
+            return;
          }
      }
      if (ret < 0) {

I think I'd do this differntly... at the beginning of the function, there is a check for kvm_check_extension(s, KVM_CAP_IRQCHIP) etc. ... I think you could now replace that check with a simple

Quoted 7 lines
	if (!kvm_has_create_irqchip) {
		return;
	}

The "kvm_vm_enable_cap(s, KVM_CAP_S390_IRQCHIP, 0)" of course has to be moved to the target/s390x/kvm/kvm.c file, too.

Sounds good. I'll do that for v2.

Thanks,
drew

#3 lore raw
In reply to Thomas Huth
Quoted 67 lines · Thomas Huth
Quoted 51 lines · Andrew Jones

Since Linux commit 00f918f61c56 ("RISC-V: KVM: Skeletal in-kernel AIA irqchip support") checking KVM_CAP_IRQCHIP returns non-zero when the RISC-V platform has AIA. The cap indicates KVM supports at least one of the following ioctls:

KVM_CREATE_IRQCHIP
KVM_IRQ_LINE
KVM_GET_IRQCHIP
KVM_SET_IRQCHIP
KVM_GET_LAPIC
KVM_SET_LAPIC

but the cap doesn't imply that KVM must support any of those ioctls in particular. However, QEMU was assuming the KVM_CREATE_IRQCHIP ioctl was supported. Stop making that assumption by introducing a KVM parameter that each architecture which supports KVM_CREATE_IRQCHIP sets. Adding parameters isn't awesome, but given how the KVM_CAP_IRQCHIP isn't very helpful on its own, we don't have a lot of options.

---

While this fixes booting guests on riscv KVM with AIA it's unlikely to get merged before the QEMU support for KVM AIA[1] lands, which would also fix the issue. I think this patch is still worth considering though since QEMU's assumption is wrong.

[1] https://lore.kernel.org/all/[email protected]/

  accel/kvm/kvm-all.c    | 5 ++++-
  include/sysemu/kvm.h   | 1 +
  target/arm/kvm.c       | 3 +++
  target/i386/kvm/kvm.c  | 2 ++
  target/s390x/kvm/kvm.c | 3 +++
  5 files changed, 13 insertions(+), 1 deletion(-)
accel/kvm/kvm-all.c +1 -0
index 373d876c0580..0f5ff8630502 100644
@@ -86,6 +86,7 @@ struct KVMParkedVcpu {
  };
  KVMState *kvm_state;
+bool kvm_has_create_irqchip;
  bool kvm_kernel_irqchip;
  bool kvm_split_irqchip;
  bool kvm_async_interrupts_allowed;
@@ -2377,8 +2378,10 @@ static void kvm_irqchip_create(KVMState *s)
          if (s->kernel_irqchip_split == ON_OFF_AUTO_ON) {

accel/kvm/kvm-all.c static void kvm_irqchip_create(KVMState *s) · line 2387

Quoted 10 lines
accel/kvm/kvm-all.c +3 -1
              error_report("Split IRQ chip mode not supported.");
              exit(1);
-        } else {
+        } else if (kvm_has_create_irqchip) {
              ret = kvm_vm_ioctl(s, KVM_CREATE_IRQCHIP);
+        } else {
+            return;
          }
      }
      if (ret < 0) {

I think I'd do this differntly... at the beginning of the function, there is a check for kvm_check_extension(s, KVM_CAP_IRQCHIP) etc. ... I think you could now replace that check with a simple

Quoted 6 lines
	if (!kvm_has_create_irqchip) {
		return;
	}

The "kvm_vm_enable_cap(s, KVM_CAP_S390_IRQCHIP, 0)" of course has to be moved to the target/s390x/kvm/kvm.c file, too.

Actually, once we've moved the s390 cap enablement to the s390 file we can just drop the whole if-else chain. We don't want the
if (!kvm_has_create_irqchip) at the top because we want to try kvm_arch_irqchip_create() even when kvm_has_create_irqchip is false, and we don't need to check KVM_CREATE_IRQCHIP for kvm_arch_irqchip_create() either. Keeping the check, as it is above in this v1, of kvm_has_create_irqchip for KVM_CREATE_IRQCHIP is still necessary, though.

Thanks,
drew