Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
Date:   Tue, 15 Aug 2023 11:34:35 -0300
From:   Jason Gunthorpe <jgg@ziepe.ca>
To:     Jie Ji <jijie.ji@linux.alibaba.com>
Cc:     dwmw2@infradead.org, baolu.lu@linux.intel.com, joro@8bytes.org,
        will@kernel.org, robin.murphy@arm.com, iommu@lists.linux.dev,
        linux-kernel@vger.kernel.org, xianting.tian@linux.alibaba.com,
        kaijieguo@linux.alibaba.com, daishengdong@yeah.net
Subject: Re: [PATCH] iommu/vt-d: Atomic breakdown of IOPT into finer
 granularity
Message-ID: <ZNuM+3R197oRfBer@ziepe.ca>
References: <20230814121016.32613-1-jijie.ji@linux.alibaba.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20230814121016.32613-1-jijie.ji@linux.alibaba.com>
Precedence: bulk

On Mon, Aug 14, 2023 at 08:10:16PM +0800, Jie Ji wrote:
> With the addition of IOMMU support for IO page fault, it's now possible
> to unpin the memory which DMA remapping. However, the lack of support
> for unmapping a subrange of the I/O page table (IOPT) in IOMMU can lead
> to some issues.
> 
> For instance, a virtual machine can establish IOPT of 2M/1G for better
> performance, while the host system enable swap and attempts to swap out
> some 4K pages. Unfortunately, unmap subrange of the large-page mapping
> will make IOMMU page walk to error level, and finally cause kernel crash.

Nothing in-kernel does something like this.

Jason