[PATCH] powerpc/book3s: Fix the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE

[PATCH] powerpc/book3s: Fix the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE

[Mahesh J Salgaonkar]

From: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>

commit id 9975f5e added new config variable CONFIG_KVM_BOOK3S_HV_POSSIBLE
that helps to select the relevant code in the kernel when HV and PR
bits are built as separate modules. As part of that commit, all the
instances of #ifdef CONFIG_KVM_BOOK3S_64_HV was replaced with
CONFIG_KVM_BOOK3S_HV_POSSIBLE. But the MCE code still depends on
CONFIG_KVM_BOOK3S_64_HV which is wrong. When HV bits are built as a
separate module the relevent MCE code gets excluded. This patch fixes
the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE.

Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
---
 arch/powerpc/kernel/exceptions-64s.S |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
index c2df815..7335857 100644
--- a/arch/powerpc/kernel/exceptions-64s.S
+++ b/arch/powerpc/kernel/exceptions-64s.S
@@ -1408,7 +1408,7 @@ machine_check_handle_early:
 	bne	9f			/* continue in V mode if we are. */
 
 5:
-#ifdef CONFIG_KVM_BOOK3S_64_HV
+#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
 	/*
 	 * We are coming from kernel context. Check if we are coming from
 	 * guest. if yes, then we can continue. We will fall through

Re: [PATCH] powerpc/book3s: Fix the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE

[Aneesh Kumar K.V]

Mahesh J Salgaonkar <mahesh@linux.vnet.ibm.com> writes:

> From: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
>
> commit id 9975f5e added new config variable CONFIG_KVM_BOOK3S_HV_POSSIBLE
> that helps to select the relevant code in the kernel when HV and PR
> bits are built as separate modules. As part of that commit, all the
> instances of #ifdef CONFIG_KVM_BOOK3S_64_HV was replaced with
> CONFIG_KVM_BOOK3S_HV_POSSIBLE. But the MCE code still depends on
> CONFIG_KVM_BOOK3S_64_HV which is wrong. When HV bits are built as a
> separate module the relevent MCE code gets excluded. This patch fixes
> the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE.
>
> Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
> ---
>  arch/powerpc/kernel/exceptions-64s.S |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
> index c2df815..7335857 100644
> --- a/arch/powerpc/kernel/exceptions-64s.S
> +++ b/arch/powerpc/kernel/exceptions-64s.S
> @@ -1408,7 +1408,7 @@ machine_check_handle_early:
>  	bne	9f			/* continue in V mode if we are. */
>
>  5:
> -#ifdef CONFIG_KVM_BOOK3S_64_HV
> +#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
>  	/*
>  	 * We are coming from kernel context. Check if we are coming from
>  	 * guest. if yes, then we can continue. We will fall through

Do we also need to update the HSTATE_IN_GUEST check to avoid checking
the PR guest mode ?ie, 

	lbz	r11,HSTATE_IN_GUEST(r13)
	cmpwi	r11,0			/* Check if coming from guest */

checking for 0 is what you want there ?

-aneesh

Re: [PATCH] powerpc/book3s: Fix the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE

[Mahesh Jagannath Salgaonkar]

On 01/07/2015 01:46 PM, Aneesh Kumar K.V wrote:
> Mahesh J Salgaonkar <mahesh@linux.vnet.ibm.com> writes:
> 
>> From: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
>>
>> commit id 9975f5e added new config variable CONFIG_KVM_BOOK3S_HV_POSSIBLE
>> that helps to select the relevant code in the kernel when HV and PR
>> bits are built as separate modules. As part of that commit, all the
>> instances of #ifdef CONFIG_KVM_BOOK3S_64_HV was replaced with
>> CONFIG_KVM_BOOK3S_HV_POSSIBLE. But the MCE code still depends on
>> CONFIG_KVM_BOOK3S_64_HV which is wrong. When HV bits are built as a
>> separate module the relevent MCE code gets excluded. This patch fixes
>> the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE.
>>
>> Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
>> ---
>>  arch/powerpc/kernel/exceptions-64s.S |    2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S
>> index c2df815..7335857 100644
>> --- a/arch/powerpc/kernel/exceptions-64s.S
>> +++ b/arch/powerpc/kernel/exceptions-64s.S
>> @@ -1408,7 +1408,7 @@ machine_check_handle_early:
>>  	bne	9f			/* continue in V mode if we are. */
>>
>>  5:
>> -#ifdef CONFIG_KVM_BOOK3S_64_HV
>> +#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
>>  	/*
>>  	 * We are coming from kernel context. Check if we are coming from
>>  	 * guest. if yes, then we can continue. We will fall through
> 
> Do we also need to update the HSTATE_IN_GUEST check to avoid checking
> the PR guest mode ?ie, 

Not required. If we are coming from guest (HSTATE_IN_GUEST != 0) we fall
through do_kvm_200->kvmppc_interrupt and for PR guest it would lead us
to kvmppc_interrupt_pr which will inject MCE in PR guest.

Having said that I just realized that I should use #ifdef
CONFIG_KVM_BOOK3S_64_HANDLER instead. The check holds good for both HV
and PR guests.

Thanks,
-Mahesh.

> 
> 	lbz	r11,HSTATE_IN_GUEST(r13)
> 	cmpwi	r11,0			/* Check if coming from guest */
> 
> checking for 0 is what you want there ?
> 
> -aneesh
> 
> _______________________________________________
> Linuxppc-dev mailing list
> Linuxppc-dev@lists.ozlabs.org
> https://lists.ozlabs.org/listinfo/linuxppc-dev
> 

Re: powerpc/book3s: Fix the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE

[Michael Ellerman]

On Wed, 2015-07-01 at 04:43:07 UTC, Mahesh Salgaonkar wrote:
> From: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
> 
> commit id 9975f5e added new config variable CONFIG_KVM_BOOK3S_HV_POSSIBLE
> that helps to select the relevant code in the kernel when HV and PR
> bits are built as separate modules. As part of that commit, all the
> instances of #ifdef CONFIG_KVM_BOOK3S_64_HV was replaced with
> CONFIG_KVM_BOOK3S_HV_POSSIBLE. But the MCE code still depends on
> CONFIG_KVM_BOOK3S_64_HV which is wrong. When HV bits are built as a
> separate module the relevent MCE code gets excluded. This patch fixes
> the MCE code to use CONFIG_KVM_BOOK3S_HV_POSSIBLE.

So what is the symptom? ie. is it fatal or just annoying.

And depending on that, should this go to stable?

cheers