On Wed, 20 Feb 2013 15:38:01 -0800, Christoffer Dall
<cdall@xxxxxxxxxxxxxxx> wrote:
> On Wed, Feb 13, 2013 at 03:47:03PM +0000, Marc Zyngier wrote:
>> Permission fault can be more than just a write fault. Execution
>> fault can also occur if the page is flagged with the XN bit.
>>
>> kvm_decode_fault() returns a value that indicates the fault type
>> (read, write, exec).
>>
>> Signed-off-by: Marc Zyngier <marc.zyngier@xxxxxxx>
>> ---
>> arch/arm/include/asm/kvm_mmu.h | 14 ++++++++++----
>> arch/arm/kvm/mmu.c | 10 ++++++----
>> 2 files changed, 16 insertions(+), 8 deletions(-)
>>
>> diff --git a/arch/arm/include/asm/kvm_mmu.h
>> b/arch/arm/include/asm/kvm_mmu.h
>> index 970f3b5..257e2bc 100644
>> --- a/arch/arm/include/asm/kvm_mmu.h
>> +++ b/arch/arm/include/asm/kvm_mmu.h
>> @@ -58,15 +58,21 @@ static inline void kvm_set_pte(pte_t *pte, pte_t
>> new_pte)
>> flush_pmd_entry(pte);
>> }
>>
>> -static inline bool kvm_is_write_fault(unsigned long hsr)
>> +enum kvm_fault_type {
>> + KVM_READ_FAULT,
>> + KVM_WRITE_FAULT,
>> + KVM_EXEC_FAULT,
>> +};
>> +
>> +static inline enum kvm_fault_type kvm_decode_fault(unsigned long hsr)
>> {
>> unsigned long hsr_ec = hsr >> HSR_EC_SHIFT;
>> if (hsr_ec == HSR_EC_IABT)
>> - return false;
>> + return KVM_EXEC_FAULT;
>> else if ((hsr & HSR_ISV) && !(hsr & HSR_WNR))
>> - return false;
>> + return KVM_READ_FAULT;
>> else
>> - return true;
>> + return KVM_WRITE_FAULT;
>> }
>>
>> static inline void kvm_clean_pgd(pgd_t *pgd)
>> diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c
>> index 633b546..58a45d1 100644
>> --- a/arch/arm/kvm/mmu.c
>> +++ b/arch/arm/kvm/mmu.c
>> @@ -510,12 +510,13 @@ static int user_mem_abort(struct kvm_vcpu *vcpu,
>> phys_addr_t fault_ipa,
>> pte_t new_pte, *ptep;
>> pfn_t pfn;
>> int ret;
>> - bool write_fault, writable;
>> + bool writable;
>> + enum kvm_fault_type fault_type;
>> unsigned long mmu_seq;
>> struct kvm_mmu_memory_cache *memcache = &vcpu->arch.mmu_page_cache;
>>
>> - write_fault = kvm_is_write_fault(kvm_vcpu_get_hsr(vcpu));
>> - if (fault_status == FSC_PERM && !write_fault) {
>> + fault_type = kvm_decode_fault(kvm_vcpu_get_hsr(vcpu));
>> + if (fault_status == FSC_PERM && fault_type == KVM_READ_FAULT) {
>> kvm_err("Unexpected L2 read permission error\n");
>> return -EFAULT;
>> }
>> @@ -537,7 +538,8 @@ static int user_mem_abort(struct kvm_vcpu *vcpu,
>> phys_addr_t fault_ipa,
>> */
>> smp_rmb();
>>
>> - pfn = gfn_to_pfn_prot(vcpu->kvm, gfn, write_fault, &writable);
>> + pfn = gfn_to_pfn_prot(vcpu->kvm, gfn,
>> + fault_type == KVM_WRITE_FAULT, &writable);
>
> This comparison within the function call looks quite ugly to me, could
> you use a variabled instead?
>
Fair enough.
M.
--
Fast, cheap, reliable. Pick two.
_______________________________________________
kvmarm mailing list
kvmarm@xxxxxxxxxxxxxxxxxxxxx
https://lists.cs.columbia.edu/cucslists/listinfo/kvmarm
