Re: [PATCH 2/2] KVM: s390: sthyi: fix specification exception detection

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 21.08.2017 14:27, Christian Borntraeger wrote:
> From: Heiko Carstens <heiko.carstens@xxxxxxxxxx>
> 
> sthyi should only generate a specification exception if the function
> code is zero and the response buffer is not on a 4k boundary.
> 
> The current code would also test for unknown function codes if the
> response buffer, that is currently only defined for function code 0,
> is not on a 4k boundary and incorrectly inject a specification
> exception instead of returning with condition code 3 and return code 4
> (unsupported function code).
> 
> Fix this by moving the boundary check.
> 
> Fixes: 95ca2cb57985 ("KVM: s390: Add sthyi emulation")
> Cc: <stable@xxxxxxxxxxxxxxx> # 4.8+
> Reviewed-by: Janosch Frank <frankja@xxxxxxxxxxxxxxxxxx>
> Signed-off-by: Heiko Carstens <heiko.carstens@xxxxxxxxxx>
> Signed-off-by: Christian Borntraeger <borntraeger@xxxxxxxxxx>
> ---
>  arch/s390/kvm/sthyi.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/s390/kvm/sthyi.c b/arch/s390/kvm/sthyi.c
> index 2773a2f..a2e5c24 100644
> --- a/arch/s390/kvm/sthyi.c
> +++ b/arch/s390/kvm/sthyi.c
> @@ -425,7 +425,7 @@ int handle_sthyi(struct kvm_vcpu *vcpu)
>  	VCPU_EVENT(vcpu, 3, "STHYI: fc: %llu addr: 0x%016llx", code, addr);
>  	trace_kvm_s390_handle_sthyi(vcpu, code, addr);
>  
> -	if (reg1 == reg2 || reg1 & 1 || reg2 & 1 || addr & ~PAGE_MASK)
> +	if (reg1 == reg2 || reg1 & 1 || reg2 & 1)
>  		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
>  
>  	if (code & 0xffff) {
> @@ -433,6 +433,9 @@ int handle_sthyi(struct kvm_vcpu *vcpu)
>  		goto out;
>  	}
>  
> +	if (addr & ~PAGE_MASK)
> +		return kvm_s390_inject_program_int(vcpu, PGM_SPECIFICATION);
> +
>  	/*
>  	 * If the page has not yet been faulted in, we want to do that
>  	 * now and not after all the expensive calculations.
> 

Reviewed-by: David Hildenbrand <david@xxxxxxxxxx>

Just wondering if this is really worth stable? (only function code 0 is
defined, so this should not happen in sane environments. or is this used
to test for support for new function codes (which would be strange but
possible)?)

-- 

Thanks,

David



[Index of Archives]     [KVM ARM]     [KVM ia64]     [KVM ppc]     [Virtualization Tools]     [Spice Development]     [Libvirt]     [Libvirt Users]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite Questions]     [Linux Kernel]     [Linux SCSI]     [XFree86]

  Powered by Linux