[Qemu-devel] [PATCH] pc: fix KVM features in pc-1.3 and earlier machine types

[Qemu-devel] [PATCH] pc: fix KVM features in pc-1.3 and earlier machine types

[Paolo Bonzini]

Due to a typo, instead of disabling KVM_FEATURE_PV_EOI (bit
6) these machine types are disabling bits 1 and 2, which are
KVM_FEATURE_NOP_IO_DELAY and KVM_FEATURE_MMU_OP.  Not a big deal
because they aren't very important and KVM_FEATURE_MMU_OP is
disabled anyway.  The worst part is actually that KVM_FEATURE_PV_EOI
is remaining enabled.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
 hw/i386/pc_piix.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/hw/i386/pc_piix.c b/hw/i386/pc_piix.c
index f0a3201..97a754e 100644
--- a/hw/i386/pc_piix.c
+++ b/hw/i386/pc_piix.c
@@ -406,7 +406,7 @@ static void pc_compat_1_3(MachineState *machine)
 static void pc_compat_1_2(MachineState *machine)
 {
     pc_compat_1_3(machine);
-    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, KVM_FEATURE_PV_EOI);
+    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, 1 << KVM_FEATURE_PV_EOI);
 }
 
 static void pc_init_pci_2_2(MachineState *machine)
@@ -483,7 +483,7 @@ static void pc_init_isa(MachineState *machine)
     if (!machine->cpu_model) {
         machine->cpu_model = "486";
     }
-    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, KVM_FEATURE_PV_EOI);
+    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, 1 << KVM_FEATURE_PV_EOI);
     enable_compat_apic_id_mode();
     pc_init1(machine, 0, 1);
 }
-- 
1.8.3.1

Re: [Qemu-devel] [PATCH] pc: fix KVM features in pc-1.3 and earlier machine types

[Eduardo Habkost]

On Mon, Jan 19, 2015 at 12:24:03PM +0100, Paolo Bonzini wrote:
> Due to a typo, instead of disabling KVM_FEATURE_PV_EOI (bit
> 6) these machine types are disabling bits 1 and 2, which are
> KVM_FEATURE_NOP_IO_DELAY and KVM_FEATURE_MMU_OP.  Not a big deal
> because they aren't very important and KVM_FEATURE_MMU_OP is
> disabled anyway.  The worst part is actually that KVM_FEATURE_PV_EOI
> is remaining enabled.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Reviewed-by: Eduardo Habkost <ehabkost@redhat.com>

-- 
Eduardo

Re: [Qemu-devel] [PATCH] pc: fix KVM features in pc-1.3 and earlier machine types

[Andreas Färber]

Am 19.01.2015 um 12:24 schrieb Paolo Bonzini:
> Due to a typo, instead of disabling KVM_FEATURE_PV_EOI (bit
> 6) these machine types are disabling bits 1 and 2, which are
> KVM_FEATURE_NOP_IO_DELAY and KVM_FEATURE_MMU_OP.  Not a big deal
> because they aren't very important and KVM_FEATURE_MMU_OP is
> disabled anyway.  The worst part is actually that KVM_FEATURE_PV_EOI
> is remaining enabled.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  hw/i386/pc_piix.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/hw/i386/pc_piix.c b/hw/i386/pc_piix.c
> index f0a3201..97a754e 100644
> --- a/hw/i386/pc_piix.c
> +++ b/hw/i386/pc_piix.c
> @@ -406,7 +406,7 @@ static void pc_compat_1_3(MachineState *machine)
>  static void pc_compat_1_2(MachineState *machine)
>  {
>      pc_compat_1_3(machine);
> -    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, KVM_FEATURE_PV_EOI);
> +    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, 1 << KVM_FEATURE_PV_EOI);
>  }
>  
>  static void pc_init_pci_2_2(MachineState *machine)
> @@ -483,7 +483,7 @@ static void pc_init_isa(MachineState *machine)
>      if (!machine->cpu_model) {
>          machine->cpu_model = "486";
>      }
> -    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, KVM_FEATURE_PV_EOI);
> +    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, 1 << KVM_FEATURE_PV_EOI);
>      enable_compat_apic_id_mode();
>      pc_init1(machine, 0, 1);
>  }

Might it make sense to tweak that helper function instead to avoid this
reoccurring? Either way, no objection, please apply through kvm queue.

Regards,
Andreas

-- 
SUSE Linux GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Felix Imendörffer, Jane Smithard, Jennifer Guild, Dilip Upmanyu,
Graham Norton; HRB 21284 (AG Nürnberg)

Re: [Qemu-devel] [PATCH] pc: fix KVM features in pc-1.3 and earlier machine types

[Paolo Bonzini]

On 19/01/2015 19:13, Andreas Färber wrote:
> Am 19.01.2015 um 12:24 schrieb Paolo Bonzini:
>> Due to a typo, instead of disabling KVM_FEATURE_PV_EOI (bit
>> 6) these machine types are disabling bits 1 and 2, which are
>> KVM_FEATURE_NOP_IO_DELAY and KVM_FEATURE_MMU_OP.  Not a big deal
>> because they aren't very important and KVM_FEATURE_MMU_OP is
>> disabled anyway.  The worst part is actually that KVM_FEATURE_PV_EOI
>> is remaining enabled.
>>
>> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
>> ---
>>  hw/i386/pc_piix.c | 4 ++--
>>  1 file changed, 2 insertions(+), 2 deletions(-)
>>
>> diff --git a/hw/i386/pc_piix.c b/hw/i386/pc_piix.c
>> index f0a3201..97a754e 100644
>> --- a/hw/i386/pc_piix.c
>> +++ b/hw/i386/pc_piix.c
>> @@ -406,7 +406,7 @@ static void pc_compat_1_3(MachineState *machine)
>>  static void pc_compat_1_2(MachineState *machine)
>>  {
>>      pc_compat_1_3(machine);
>> -    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, KVM_FEATURE_PV_EOI);
>> +    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, 1 << KVM_FEATURE_PV_EOI);
>>  }
>>  
>>  static void pc_init_pci_2_2(MachineState *machine)
>> @@ -483,7 +483,7 @@ static void pc_init_isa(MachineState *machine)
>>      if (!machine->cpu_model) {
>>          machine->cpu_model = "486";
>>      }
>> -    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, KVM_FEATURE_PV_EOI);
>> +    x86_cpu_compat_kvm_no_autoenable(FEAT_KVM, 1 << KVM_FEATURE_PV_EOI);
>>      enable_compat_apic_id_mode();
>>      pc_init1(machine, 0, 1);
>>  }
> 
> Might it make sense to tweak that helper function instead to avoid this
> reoccurring? Either way, no objection, please apply through kvm queue.

Will do.  The problem is that the same function is used for CPUID_*
defines that _are_ bitmasks.

Paolo