Received-SPF: pass (google.com: domain of linux-kernel+bounces-65363-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:40f1:3f00::1 as permitted sender) client-ip=2604:1380:40f1:3f00::1;
Feedback-ID: ie3994620:Fastmail
Date: Wed, 14 Feb 2024 16:49:43 +0200
From: "Kirill A. Shutemov" <kirill@shutemov.name>
To: Vishal Annapurve <vannapurve@google.com>
Cc: x86@kernel.org, linux-kernel@vger.kernel.org, pbonzini@redhat.com, 
	rientjes@google.com, bgardon@google.com, seanjc@google.com, erdemaktas@google.com, 
	ackerleytng@google.com, jxgao@google.com, sagis@google.com, oupton@google.com, 
	peterx@redhat.com, vkuznets@redhat.com, dmatlack@google.com, pgonda@google.com, 
	michael.roth@amd.com, thomas.lendacky@amd.com, dave.hansen@linux.intel.com, 
	linux-coco@lists.linux.dev, chao.p.peng@linux.intel.com, isaku.yamahata@gmail.com, 
	andrew.jones@linux.dev, corbet@lwn.net, hch@lst.de, m.szyprowski@samsung.com, bp@suse.de, 
	rostedt@goodmis.org, iommu@lists.linux.dev
Subject: Re: [RFC V1 1/5] swiotlb: Support allocating DMA memory from SWIOTLB
Message-ID: <etvf2mon464whscbxqktdd7bputnqmmwmoeg7ssixsk4kljfek@4wngbgzbbmck>
References: <20240112055251.36101-1-vannapurve@google.com>
 <20240112055251.36101-2-vannapurve@google.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20240112055251.36101-2-vannapurve@google.com>

On Fri, Jan 12, 2024 at 05:52:47AM +0000, Vishal Annapurve wrote:
> Modify SWIOTLB framework to allocate DMA memory always from SWIOTLB.
> 
> CVMs use SWIOTLB buffers for bouncing memory when using dma_map_* APIs
> to setup memory for IO operations. SWIOTLB buffers are marked as shared
> once during early boot.
> 
> Buffers allocated using dma_alloc_* APIs are allocated from kernel memory
> and then converted to shared during each API invocation. This patch ensures
> that such buffers are also allocated from already shared SWIOTLB
> regions. This allows enforcing alignment requirements on regions marked
> as shared.

But does it work in practice? 

Some devices (like GPUs) require a lot of DMA memory. So with this approach
we would need to have huge SWIOTLB buffer that is unused in most VMs.

-- 
  Kiryl Shutsemau / Kirill A. Shutemov