[kernel] KVM: PPC: Preserve storage control bits

PR KVM page fault handler performs eaddr to pte translation for a guest,
however kvmppc_mmu_book3s_64_xlate() does not preserve WIMG bits
(storage control) in the kvmppc_pte struct. If PR KVM is running as
a second level guest under HV KVM, and PR KVM tries inserting HPT entry,
this fails in HV KVM if it already has this mapping.

This preserves WIMG bits between kvmppc_mmu_book3s_64_xlate() and
kvmppc_mmu_map_page().

Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
---

This allows MMIO BAR mapping for nested guest with VFIO.

This is the check in HV KVM which failed:

arch/powerpc/kvm/book3s_hv_rm_mmu.c
long kvmppc_do_h_enter(struct kvm *kvm, unsigned long flags,
[...]

        /*If we had host pte mapping then  Check WIMG */
        if (ptep && !hpte_cache_flags_ok(ptel, is_ci)) {
                if (is_ci)
                        return H_PARAMETER;
                /*
                 * Allow guest to map emulated device memory as
                 * uncacheable, but actually make it cacheable.
                 */
                ptel &= ~(HPTE_R_W|HPTE_R_I|HPTE_R_G);
                ptel |= HPTE_R_M;
        }
---
 arch/powerpc/include/asm/kvm_host.h   | 1 +
 arch/powerpc/kvm/book3s_64_mmu.c      | 1 +
 arch/powerpc/kvm/book3s_64_mmu_host.c | 2 ++
 arch/powerpc/kvm/book3s_pr.c          | 2 +-
 4 files changed, 5 insertions(+), 1 deletion(-)

On Fri, Mar 24, 2017 at 05:49:22PM +1100, Alexey Kardashevskiy wrote:
> PR KVM page fault handler performs eaddr to pte translation for a guest,
> however kvmppc_mmu_book3s_64_xlate() does not preserve WIMG bits
> (storage control) in the kvmppc_pte struct. If PR KVM is running as
> a second level guest under HV KVM, and PR KVM tries inserting HPT entry,
> this fails in HV KVM if it already has this mapping.
> 
> This preserves WIMG bits between kvmppc_mmu_book3s_64_xlate() and
> kvmppc_mmu_map_page().
> 
> Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>

Reviewed-by: David Gibson <david@gibson.dropbear.id.au>

> ---
> 
> This allows MMIO BAR mapping for nested guest with VFIO.
> 
> This is the check in HV KVM which failed:
> 
> arch/powerpc/kvm/book3s_hv_rm_mmu.c
> long kvmppc_do_h_enter(struct kvm *kvm, unsigned long flags,
> [...]
> 
>         /*If we had host pte mapping then  Check WIMG */
>         if (ptep && !hpte_cache_flags_ok(ptel, is_ci)) {
>                 if (is_ci)
>                         return H_PARAMETER;
>                 /*
>                  * Allow guest to map emulated device memory as
>                  * uncacheable, but actually make it cacheable.
>                  */
>                 ptel &= ~(HPTE_R_W|HPTE_R_I|HPTE_R_G);
>                 ptel |= HPTE_R_M;
>         }
> ---
>  arch/powerpc/include/asm/kvm_host.h   | 1 +
>  arch/powerpc/kvm/book3s_64_mmu.c      | 1 +
>  arch/powerpc/kvm/book3s_64_mmu_host.c | 2 ++
>  arch/powerpc/kvm/book3s_pr.c          | 2 +-
>  4 files changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/powerpc/include/asm/kvm_host.h b/arch/powerpc/include/asm/kvm_host.h
> index 7bba8f415627..bf6822cd4f86 100644
> --- a/arch/powerpc/include/asm/kvm_host.h
> +++ b/arch/powerpc/include/asm/kvm_host.h
> @@ -345,6 +345,7 @@ struct kvmppc_pte {
>  	bool may_read		: 1;
>  	bool may_write		: 1;
>  	bool may_execute	: 1;
> +	unsigned long wimg;
>  	u8 page_size;		/* MMU_PAGE_xxx */
>  };
>  
> diff --git a/arch/powerpc/kvm/book3s_64_mmu.c b/arch/powerpc/kvm/book3s_64_mmu.c
> index 70153578131a..29ebe2fd5867 100644
> --- a/arch/powerpc/kvm/book3s_64_mmu.c
> +++ b/arch/powerpc/kvm/book3s_64_mmu.c
> @@ -319,6 +319,7 @@ static int kvmppc_mmu_book3s_64_xlate(struct kvm_vcpu *vcpu, gva_t eaddr,
>  		gpte->may_execute = true;
>  	gpte->may_read = false;
>  	gpte->may_write = false;
> +	gpte->wimg = r & HPTE_R_WIMG;
>  
>  	switch (pp) {
>  	case 0:
> diff --git a/arch/powerpc/kvm/book3s_64_mmu_host.c b/arch/powerpc/kvm/book3s_64_mmu_host.c
> index 4b4e927c4822..145a61892c48 100644
> --- a/arch/powerpc/kvm/book3s_64_mmu_host.c
> +++ b/arch/powerpc/kvm/book3s_64_mmu_host.c
> @@ -145,6 +145,8 @@ int kvmppc_mmu_map_page(struct kvm_vcpu *vcpu, struct kvmppc_pte *orig_pte,
>  	else
>  		kvmppc_mmu_flush_icache(pfn);
>  
> +	rflags = (rflags & ~HPTE_R_WIMG) | orig_pte->wimg;
> +
>  	/*
>  	 * Use 64K pages if possible; otherwise, on 64K page kernels,
>  	 * we need to transfer 4 more bits from guest real to host real addr.
> diff --git a/arch/powerpc/kvm/book3s_pr.c b/arch/powerpc/kvm/book3s_pr.c
> index ce437b98477e..f026b062c0ed 100644
> --- a/arch/powerpc/kvm/book3s_pr.c
> +++ b/arch/powerpc/kvm/book3s_pr.c
> @@ -537,7 +537,7 @@ int kvmppc_handle_pagefault(struct kvm_run *run, struct kvm_vcpu *vcpu,
>  	int r = RESUME_GUEST;
>  	int relocated;
>  	int page_found = 0;
> -	struct kvmppc_pte pte;
> +	struct kvmppc_pte pte = { 0 };
>  	bool dr = (kvmppc_get_msr(vcpu) & MSR_DR) ? true : false;
>  	bool ir = (kvmppc_get_msr(vcpu) & MSR_IR) ? true : false;
>  	u64 vsid;