Re: [PATCH 4/6] KVM: arm64/mmu: avoid struct page in MMU

From: Marc Zyngier <maz@kernel.org>
To: David Stevens <stevensd@chromium.org>
Cc: Huacai Chen <chenhuacai@kernel.org>,
	Aleksandar Markovic <aleksandar.qemu.devel@gmail.com>,
	Paul Mackerras <paulus@ozlabs.org>,
	Paolo Bonzini <pbonzini@redhat.com>,
	Zhenyu Wang <zhenyuw@linux.intel.com>,
	Zhi Wang <zhi.a.wang@intel.com>,
	James Morse <james.morse@arm.com>,
	Alexandru Elisei <alexandru.elisei@arm.com>,
	Suzuki K Poulose <suzuki.poulose@arm.com>,
	Will Deacon <will@kernel.org>,
	Sean Christopherson <seanjc@google.com>,
	Vitaly Kuznetsov <vkuznets@redhat.com>,
	Wanpeng Li <wanpengli@tencent.com>,
	Jim Mattson <jmattson@google.com>, Joerg Roedel <joro@8bytes.org>,
	linux-arm-kernel@lists.infradead.org,
	kvmarm@lists.cs.columbia.edu, linux-kernel@vger.kernel.org,
	linux-mips@vger.kernel.org, kvm@vger.kernel.org,
	kvm-ppc@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
	intel-gvt-dev@lists.freedesktop.org,
	intel-gfx@lists.freedesktop.org, dri-devel@lists.freedesktop.org
Subject: Re: [PATCH 4/6] KVM: arm64/mmu: avoid struct page in MMU
Date: Thu, 24 Jun 2021 11:43:19 +0100	[thread overview]
Message-ID: <87lf6zilvc.wl-maz@kernel.org> (raw)
In-Reply-To: <20210624035749.4054934-5-stevensd@google.com>

On Thu, 24 Jun 2021 04:57:47 +0100,
David Stevens <stevensd@chromium.org> wrote:
> 
> From: David Stevens <stevensd@chromium.org>
> 
> Avoid converting pfns returned by follow_fault_pfn to struct pages to
> transiently take a reference. The reference was originally taken to
> match the reference taken by gup. However, pfns returned by
> follow_fault_pfn may not have a struct page set up for reference
> counting.
> 
> Signed-off-by: David Stevens <stevensd@chromium.org>
> ---
>  arch/arm64/kvm/mmu.c | 43 +++++++++++++++++++++++--------------------
>  1 file changed, 23 insertions(+), 20 deletions(-)
> 
> diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c
> index 896b3644b36f..a741972cb75f 100644
> --- a/arch/arm64/kvm/mmu.c
> +++ b/arch/arm64/kvm/mmu.c

[...]

> @@ -968,16 +968,16 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
>  	 */
>  	if (vma_pagesize == PAGE_SIZE && !force_pte)
>  		vma_pagesize = transparent_hugepage_adjust(memslot, hva,
> -							   &pfn, &fault_ipa);
> +							   &pfnpg, &fault_ipa);
>  	if (writable)
>  		prot |= KVM_PGTABLE_PROT_W;
>  
>  	if (fault_status != FSC_PERM && !device)
> -		clean_dcache_guest_page(pfn, vma_pagesize);
> +		clean_dcache_guest_page(pfnpg.pfn, vma_pagesize);
>  
>  	if (exec_fault) {
>  		prot |= KVM_PGTABLE_PROT_X;
> -		invalidate_icache_guest_page(pfn, vma_pagesize);
> +		invalidate_icache_guest_page(pfnpg.pfn, vma_pagesize);

This is going to clash with what is currently in -next, specially with
MTE.

Paolo, if you really want to take this in 5.13, can you please push a
stable branch based on -rc4 or older so that I can merge it in and
test it?

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.