Linux PCI subsystem development
 help / color / mirror / Atom feed
From: Nathan Lynch via B4 Relay <devnull+nathan.lynch.amd.com@kernel.org>
To: Vinod Koul <vkoul@kernel.org>
Cc: Wei Huang <wei.huang2@amd.com>,
	 Mario Limonciello <mario.limonciello@amd.com>,
	 Bjorn Helgaas <bhelgaas@google.com>,
	linux-pci@vger.kernel.org,  linux-kernel@vger.kernel.org,
	dmaengine@vger.kernel.org
Subject: [PATCH RFC 00/13] dmaengine: Smart Data Accelerator Interface (SDXI) basic support
Date: Fri, 05 Sep 2025 13:48:23 -0500	[thread overview]
Message-ID: <20250905-sdxi-base-v1-0-d0341a1292ba@amd.com> (raw)

The Smart Data Accelerator Interface (SDXI) is a vendor-neutral
architecture for memory-to-memory data movement offload designed for
kernel bypass and virtualization.

General information on SDXI may be found at:
https://www.snia.org/sdxi

This submission adds a driver with basic support for PCIe-hosted SDXI 1.0
implementations and includes a DMA engine provider.

It is very much a work in progress. Among other issues, the DMA
provider code only supports single-threaded polled mode, and context
management should use better data structures.

While we're addressing those shortcomings, we'd appreciate any
feedback on:

* Where the code should live. SDXI entails a fair amount of code for
  context and descriptor management, and we expect to eventually add a
  character device ABI for user space access. Should all of this go in
  drivers/dma/sdxi?

* Whether the DMA engine provider should use virt-dma/vchan. SDXI
  submission queues can be almost arbitrarily large, and I'm not sure
  putting a software queue in front of that makes sense.

Planned future SDXI work (out of scope for this series):

* Character device for user space access. We are evaluating the uacce
  framework for this.

* Support for operation types to be added in future SDXI revisions.

* Greater configurability for control structures, e.g. descriptor ring
  size.

The latest released version of the SDXI specification is 1.0:
https://www.snia.org/sites/default/files/technical-work/sdxi/release/SNIA-SDXI-Specification-v1.0a.pdf

Draft versions of future SDXI specifications in development may be found at:
https://www.snia.org/tech_activities/publicreview#sdxi

---
Nathan Lynch (13):
      PCI: Add SNIA SDXI accelerator sub-class
      dmaengine: sdxi: Add control structure definitions
      dmaengine: sdxi: Add descriptor encoding and unit tests
      dmaengine: sdxi: Add MMIO register definitions
      dmaengine: sdxi: Add software data structures
      dmaengine: sdxi: Add error reporting support
      dmaengine: sdxi: Import descriptor enqueue code from spec
      dmaengine: sdxi: Context creation/removal, descriptor submission
      dmaengine: sdxi: Add core device management code
      dmaengine: sdxi: Add PCI driver support
      dmaengine: sdxi: Add DMA engine provider
      dmaengine: sdxi: Add Kconfig and Makefile
      MAINTAINERS: Add entry for SDXI driver

 MAINTAINERS                         |   7 +
 drivers/dma/Kconfig                 |   2 +
 drivers/dma/Makefile                |   1 +
 drivers/dma/sdxi/.kunitconfig       |   4 +
 drivers/dma/sdxi/Kconfig            |  23 ++
 drivers/dma/sdxi/Makefile           |  17 ++
 drivers/dma/sdxi/context.c          | 547 ++++++++++++++++++++++++++++++++++++
 drivers/dma/sdxi/context.h          |  28 ++
 drivers/dma/sdxi/descriptor.c       | 197 +++++++++++++
 drivers/dma/sdxi/descriptor.h       | 107 +++++++
 drivers/dma/sdxi/descriptor_kunit.c | 181 ++++++++++++
 drivers/dma/sdxi/device.c           | 401 ++++++++++++++++++++++++++
 drivers/dma/sdxi/dma.c              | 409 +++++++++++++++++++++++++++
 drivers/dma/sdxi/dma.h              |  12 +
 drivers/dma/sdxi/enqueue.c          | 136 +++++++++
 drivers/dma/sdxi/enqueue.h          |  16 ++
 drivers/dma/sdxi/error.c            | 340 ++++++++++++++++++++++
 drivers/dma/sdxi/error.h            |  16 ++
 drivers/dma/sdxi/hw.h               | 249 ++++++++++++++++
 drivers/dma/sdxi/mmio.h             |  92 ++++++
 drivers/dma/sdxi/pci.c              | 216 ++++++++++++++
 drivers/dma/sdxi/sdxi.h             | 206 ++++++++++++++
 include/linux/pci_ids.h             |   1 +
 23 files changed, 3208 insertions(+)
---
base-commit: c17b750b3ad9f45f2b6f7e6f7f4679844244f0b9
change-id: 20250813-sdxi-base-73d7c9fdce57

Best regards,
-- 
Nathan Lynch <nathan.lynch@amd.com>



             reply	other threads:[~2025-09-05 18:48 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-09-05 18:48 Nathan Lynch via B4 Relay [this message]
2025-09-05 18:48 ` [PATCH RFC 01/13] PCI: Add SNIA SDXI accelerator sub-class Nathan Lynch via B4 Relay
2025-09-15 17:25   ` Bjorn Helgaas
2025-09-15 20:17     ` Nathan Lynch
2025-09-05 18:48 ` [PATCH RFC 02/13] dmaengine: sdxi: Add control structure definitions Nathan Lynch via B4 Relay
2025-09-05 18:48 ` [PATCH RFC 03/13] dmaengine: sdxi: Add descriptor encoding and unit tests Nathan Lynch via B4 Relay
2025-09-15 11:52   ` Jonathan Cameron
2025-09-15 19:30     ` Nathan Lynch
2025-09-16 14:20       ` Jonathan Cameron
2025-09-16 19:06         ` Nathan Lynch
2025-09-05 18:48 ` [PATCH RFC 04/13] dmaengine: sdxi: Add MMIO register definitions Nathan Lynch via B4 Relay
2025-09-05 18:48 ` [PATCH RFC 05/13] dmaengine: sdxi: Add software data structures Nathan Lynch via B4 Relay
2025-09-15 11:59   ` Jonathan Cameron
2025-09-16 19:07     ` Nathan Lynch
2025-09-16  9:38   ` Markus Elfring
2025-09-05 18:48 ` [PATCH RFC 06/13] dmaengine: sdxi: Add error reporting support Nathan Lynch via B4 Relay
2025-09-15 12:11   ` Jonathan Cameron
2025-09-15 20:42     ` Nathan Lynch
2025-09-16 14:23       ` Jonathan Cameron
2025-09-05 18:48 ` [PATCH RFC 07/13] dmaengine: sdxi: Import descriptor enqueue code from spec Nathan Lynch via B4 Relay
2025-09-15 12:18   ` Jonathan Cameron
2025-09-16 17:05   ` [External] : " ALOK TIWARI
2025-09-05 18:48 ` [PATCH RFC 08/13] dmaengine: sdxi: Context creation/removal, descriptor submission Nathan Lynch via B4 Relay
2025-09-15 14:12   ` Jonathan Cameron
2025-09-16 20:40     ` Nathan Lynch
2025-09-17 13:34       ` Jonathan Cameron
2025-09-15 19:42   ` Markus Elfring
2025-09-05 18:48 ` [PATCH RFC 09/13] dmaengine: sdxi: Add core device management code Nathan Lynch via B4 Relay
2025-09-15 14:23   ` Jonathan Cameron
2025-09-16 21:23     ` Nathan Lynch
2025-09-05 18:48 ` [PATCH RFC 10/13] dmaengine: sdxi: Add PCI driver support Nathan Lynch via B4 Relay
2025-09-05 19:14   ` Mario Limonciello
2025-09-10 15:25     ` Nathan Lynch
2025-09-05 20:05   ` Bjorn Helgaas
2025-09-10 15:28     ` Nathan Lynch
2025-09-15 15:03   ` Jonathan Cameron
2025-09-16 16:43   ` [External] : " ALOK TIWARI
2025-09-05 18:48 ` [PATCH RFC 11/13] dmaengine: sdxi: Add DMA engine provider Nathan Lynch via B4 Relay
2025-09-15 15:16   ` Jonathan Cameron
2025-09-05 18:48 ` [PATCH RFC 12/13] dmaengine: sdxi: Add Kconfig and Makefile Nathan Lynch via B4 Relay
2025-09-15 15:08   ` Jonathan Cameron
2025-09-15 16:44     ` Nathan Lynch
2025-09-05 18:48 ` [PATCH RFC 13/13] MAINTAINERS: Add entry for SDXI driver Nathan Lynch via B4 Relay

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250905-sdxi-base-v1-0-d0341a1292ba@amd.com \
    --to=devnull+nathan.lynch.amd.com@kernel.org \
    --cc=bhelgaas@google.com \
    --cc=dmaengine@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=mario.limonciello@amd.com \
    --cc=nathan.lynch@amd.com \
    --cc=vkoul@kernel.org \
    --cc=wei.huang2@amd.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox