Sender: Paolo Bonzini <paolo.bonzini@gmail.com>
Message-ID: <5f6f43a1-bb4c-f498-2aba-4c93ab57fc98@gnu.org>
Date:   Tue, 23 Nov 2021 09:54:52 +0100
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.2.0
Subject: Re: [RFC v2 PATCH 01/13] mm/shmem: Introduce F_SEAL_GUEST
Content-Language: en-US
To:     Chao Peng <chao.p.peng@linux.intel.com>, kvm@vger.kernel.org,
        linux-kernel@vger.kernel.org, linux-mm@kvack.org,
        linux-fsdevel@vger.kernel.org, qemu-devel@nongnu.org
Cc:     Jonathan Corbet <corbet@lwn.net>,
        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>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        x86@kernel.org, "H . Peter Anvin" <hpa@zytor.com>,
        Hugh Dickins <hughd@google.com>,
        Jeff Layton <jlayton@kernel.org>,
        "J . Bruce Fields" <bfields@fieldses.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        Yu Zhang <yu.c.zhang@linux.intel.com>,
        "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>,
        luto@kernel.org, john.ji@intel.com, susie.li@intel.com,
        jun.nakajima@intel.com, dave.hansen@intel.com, ak@linux.intel.com,
        david@redhat.com
References: <20211119134739.20218-1-chao.p.peng@linux.intel.com>
 <20211119134739.20218-2-chao.p.peng@linux.intel.com>
From:   Paolo Bonzini <bonzini@gnu.org>
In-Reply-To: <20211119134739.20218-2-chao.p.peng@linux.intel.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 11/19/21 14:47, Chao Peng wrote:
> +static void guest_invalidate_page(struct inode *inode,
> +				  struct page *page, pgoff_t start, pgoff_t end)
> +{
> +	struct shmem_inode_info *info = SHMEM_I(inode);
> +
> +	if (!info->guest_ops || !info->guest_ops->invalidate_page_range)
> +		return;
> +
> +	start = max(start, page->index);
> +	end = min(end, page->index + thp_nr_pages(page)) - 1;
> +
> +	info->guest_ops->invalidate_page_range(inode, info->guest_owner,
> +					       start, end);
> +}

The lack of protection makes the API quite awkward to use;
the usual way to do this is with refcount_inc_not_zero (aka 
kvm_get_kvm_safe).

Can you use the shmem_inode_info spinlock to protect against this?  If 
register/unregister take the spinlock, the invalidate and fallocate can 
take a reference under the same spinlock, like this:

	if (!info->guest_ops)
		return;

	spin_lock(&info->lock);
	ops = info->guest_ops;
	if (!ops) {
		spin_unlock(&info->lock);
		return;
	}

	/* Calls kvm_get_kvm_safe.  */
	r = ops->get_guest_owner(info->guest_owner);
	spin_unlock(&info->lock);
	if (r < 0)
		return;

	start = max(start, page->index);
	end = min(end, page->index + thp_nr_pages(page)) - 1;

	ops->invalidate_page_range(inode, info->guest_owner,
					       start, end);
	ops->put_guest_owner(info->guest_owner);

Considering that you have to take a mutex anyway in patch 13, and that 
the critical section here is very small, the extra indirect calls are 
cheaper than walking the vm_list; and it makes the API clearer.

Paolo