[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CACzX3Au7PW2zFFLmtNgW10wq+Kp-bp66GXUVCUCfS4VvK3tDYw@mail.gmail.com>
Date: Mon, 24 Nov 2025 19:05:54 +0530
From: Anuj gupta <anuj1072538@...il.com>
To: Pavel Begunkov <asml.silence@...il.com>
Cc: linux-block@...r.kernel.org, io-uring@...r.kernel.org,
Vishal Verma <vishal1.verma@...el.com>, tushar.gohad@...el.com,
Keith Busch <kbusch@...nel.org>, Jens Axboe <axboe@...nel.dk>, Christoph Hellwig <hch@....de>,
Sagi Grimberg <sagi@...mberg.me>, Alexander Viro <viro@...iv.linux.org.uk>,
Christian Brauner <brauner@...nel.org>, Andrew Morton <akpm@...ux-foundation.org>,
Sumit Semwal <sumit.semwal@...aro.org>, Christian König <christian.koenig@....com>,
linux-kernel@...r.kernel.org, linux-nvme@...ts.infradead.org,
linux-fsdevel@...r.kernel.org, linux-media@...r.kernel.org,
dri-devel@...ts.freedesktop.org, linaro-mm-sig@...ts.linaro.org
Subject: Re: [RFC v2 00/11] Add dmabuf read/write via io_uring
This series significantly reduces the IOMMU/DMA overhead for I/O,
particularly when the IOMMU is configured in STRICT or LAZY mode. I
modified t/io_uring in fio to exercise this path and tested with an
Intel Optane device. On my setup, I see the following improvement:
- STRICT: before = 570 KIOPS, after = 5.01 MIOPS
- LAZY: before = 1.93 MIOPS, after = 5.01 MIOPS
- PASSTHROUGH: before = 5.01 MIOPS, after = 5.01 MIOPS
The STRICT/LAZY numbers clearly show the benefit of avoiding per-I/O
dma_map/dma_unmap and reusing the pre-mapped DMA addresses.
--
Anuj Gupta
Powered by blists - more mailing lists