<feed xmlns='http://www.w3.org/2005/Atom'>
<title>wireguard-linux/Documentation/driver-api/dmaengine, branch stable</title>
<subtitle>WireGuard for the Linux kernel</subtitle>
<id>https://git.zx2c4.com/wireguard-linux/atom/Documentation/driver-api/dmaengine?h=stable</id>
<link rel='self' href='https://git.zx2c4.com/wireguard-linux/atom/Documentation/driver-api/dmaengine?h=stable'/>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/'/>
<updated>2026-10-11T20:32:55Z</updated>
<entry>
<title>docs: dmaengine: clarify ordering rules around cookie completion</title>
<updated>2026-10-11T20:32:55Z</updated>
<author>
<name>Vinod Koul</name>
<email>vkoul@kernel.org</email>
</author>
<published>2026-10-09T09:37:44Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=fa89be96e845de705eed6edd01ecb79d46ee21c7'/>
<id>urn:sha1:fa89be96e845de705eed6edd01ecb79d46ee21c7</id>
<content type='text'>
Sashiko pointed in [1] about the cookie completion order. It is good to
Document that for ensuring that dma buffer is flushed and no stale data
is seeing by consumer, we need to make sure such action is taken before
drivers invoke dma cookie completion.

[1]: https://lore.kernel.org/r/20260917072507.5BDB61F000FF@smtp.kernel.org

Reviewed-by: Frank Li &lt;Frank.Li@nxp.com&gt;
Reviewed-by: Dave Jiang &lt;dave.jiang@intel.com&gt;
Reviewed-by: Wolfram Sang &lt;wsa+renesas@sang-engineering.com&gt;
Link: https://patch.msgid.link/20261009093745.539924-1-vkoul@kernel.org
Signed-off-by: Vinod Koul &lt;vkoul@kernel.org&gt;
</content>
</entry>
<entry>
<title>docs: dmaengine: document ownership of the descriptor flags field</title>
<updated>2026-10-05T14:12:30Z</updated>
<author>
<name>Bogdan Mahomet</name>
<email>bogdan.mahomet.dev@gmail.com</email>
</author>
<published>2026-08-22T00:15:30Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=ef399305a5bf64231059c63a53f3384eb59c2458'/>
<id>urn:sha1:ef399305a5bf64231059c63a53f3384eb59c2458</id>
<content type='text'>
The flags field of struct dma_async_tx_descriptor has carried a TODO
since the provider documentation was added in 2014, asking whether a
driver may modify the field itself or whether it must always hold the
flags passed in the arguments.

The field belongs to the client. dma_async_tx_descriptor_init() does
not touch it, so the provider has to set it, and providers store the
flags they were given verbatim; vchan_tx_prep() does this for every
driver built on top of virt-dma. Afterwards providers only read the
field: the DMA_PREP_* bits are consumed while the descriptor is being
prepared, and DMA_CTRL_ACK and DMA_CTRL_REUSE are tested with
async_tx_test_ack() and dmaengine_desc_test_reuse() before a
descriptor is recycled or freed.

Once a descriptor has been prepared, nothing in the tree changes the
field except the client helpers in include/linux/dmaengine.h:
async_tx_ack(), async_tx_clear_ack(), dmaengine_desc_set_reuse() and
dmaengine_desc_clear_reuse().

Document that rule in place of the TODO.

Signed-off-by: Bogdan Mahomet &lt;bogdan.mahomet.dev@gmail.com&gt;
Reviewed-by: Frank Li &lt;Frank.Li@nxp.com&gt;
Link: https://patch.msgid.link/20260822001536.29269-1-bogdan.mahomet.dev@gmail.com
Signed-off-by: Vinod Koul &lt;vkoul@kernel.org&gt;
</content>
</entry>
<entry>
<title>docs: dmaengine: Fix bracket</title>
<updated>2026-09-03T13:32:13Z</updated>
<author>
<name>Manuel Ebner</name>
<email>manuelebnerli@mailbox.org</email>
</author>
<published>2026-09-03T09:55:45Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=635e425c0771ec2ba09b0ef69bb9814e2554374e'/>
<id>urn:sha1:635e425c0771ec2ba09b0ef69bb9814e2554374e</id>
<content type='text'>
Add missing ')' to sentence.

Signed-off-by: Manuel Ebner &lt;manuelebnerli@mailbox.org&gt;
Link: https://patch.msgid.link/20260903095545.729607-3-manuelebnerli@mailbox.org
Signed-off-by: Vinod Koul &lt;vkoul@kernel.org&gt;
</content>
</entry>
<entry>
<title>dmaengine: Add API to combine configuration and preparation (sg and single)</title>
<updated>2026-06-30T11:12:30Z</updated>
<author>
<name>Frank Li</name>
<email>Frank.Li@nxp.com</email>
</author>
<published>2026-05-21T15:32:47Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=796bdb33e86aec8504bf8868e0665f120638ac72'/>
<id>urn:sha1:796bdb33e86aec8504bf8868e0665f120638ac72</id>
<content type='text'>
Previously, configuration and preparation required two separate calls. This
works well when configuration is done only once during initialization.

However, in cases where the burst length or source/destination address must
be adjusted for each transfer, calling two functions is verbose and
requires additional locking to ensure both steps complete atomically.

Add a new API dmaengine_prep_config_single() and dmaengine_prep_config_sg()
and callback device_prep_config_sg() that combines configuration and
preparation into a single operation. If the configuration argument is
passed as NULL, fall back to the existing implementation.

Tested-by: Niklas Cassel &lt;cassel@kernel.org&gt;
Acked-by: Manivannan Sadhasivam &lt;mani@kernel.org&gt;
Signed-off-by: Frank Li &lt;Frank.Li@nxp.com&gt;
Link: https://patch.msgid.link/20260521-dma_prep_config-v7-1-1f73f4899883@nxp.com
Signed-off-by: Vinod Koul &lt;vkoul@kernel.org&gt;
</content>
</entry>
<entry>
<title>docs: driver-api: fix 6 spelling typos in Documentation/driver-api</title>
<updated>2026-03-25T18:36:02Z</updated>
<author>
<name>Tomás Pando</name>
<email>tovictakamine@gmail.com</email>
</author>
<published>2026-03-24T16:36:04Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=19dcccbc064d6c58eaafae1ecb94821a2535cc26'/>
<id>urn:sha1:19dcccbc064d6c58eaafae1ecb94821a2535cc26</id>
<content type='text'>
Fix minor spelling mistakes in the driver-api documentation. These
changes improve readability in ACPI, CXL, DMA and PCI docs.

Reviewed-by: Randy Dunlap &lt;rdunlap@infradead.org&gt;
Signed-off-by: Tomás Pando &lt;tovictakamine@gmail.com&gt;
Signed-off-by: Jonathan Corbet &lt;corbet@lwn.net&gt;
Message-ID: &lt;20260324163604.5710-1-tovictakamine@gmail.com&gt;
</content>
</entry>
<entry>
<title>Merge tag 'dmaengine-7.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine</title>
<updated>2026-02-17T19:47:17Z</updated>
<author>
<name>Linus Torvalds</name>
<email>torvalds@linux-foundation.org</email>
</author>
<published>2026-02-17T19:47:17Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=e81dd54f62c753dd423d1a9b62481a1c599fb975'/>
<id>urn:sha1:e81dd54f62c753dd423d1a9b62481a1c599fb975</id>
<content type='text'>
Pull dmaengine updates from Vinod Koul:
 "Core:
   - Add Frank Li as susbstem reviewer to help with reviews

  New Support:
   - Mediatek support for Dimensity 6300 and 9200 controller
   - Qualcomm Kaanapali and Glymur GPI DMA engine
   - Synopsis DW AXI Agilex5
   - Renesas RZ/V2N SoC
   - Atmel microchip lan9691-dma
   - Tegra ADMA tegra264

  Updates:
   - sg_nents_for_dma() helper use in subsystem
   - pm_runtime_mark_last_busy() redundant call update for subsystem
   - Residue support for xilinx AXIDMA driver
   - Intel Max SGL Size Support and capabilities for DSA3.0
   - AXI dma larger than 32bits address support"

* tag 'dmaengine-7.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine: (64 commits)
  dmaengine: add Frank Li as reviewer
  dt-bindings: dma: qcom,gpi: Update max interrupts lines to 16
  dmaengine: fsl-edma: don't explicitly disable clocks in .remove()
  dmaengine: xilinx: xdma: use sg_nents_for_dma() helper
  dmaengine: sh: use sg_nents_for_dma() helper
  dmaengine: sa11x0: use sg_nents_for_dma() helper
  dmaengine: qcom: bam_dma: use sg_nents_for_dma() helper
  dmaengine: qcom: adm: use sg_nents_for_dma() helper
  dmaengine: pxa-dma: use sg_nents_for_dma() helper
  dmaengine: lgm: use sg_nents_for_dma() helper
  dmaengine: k3dma: use sg_nents_for_dma() helper
  dmaengine: dw-axi-dmac: use sg_nents_for_dma() helper
  dmaengine: bcm2835-dma: use sg_nents_for_dma() helper
  dmaengine: axi-dmac: use sg_nents_for_dma() helper
  dmaengine: altera-msgdma: use sg_nents_for_dma() helper
  scatterlist: introduce sg_nents_for_dma() helper
  dmaengine: idxd: Add Max SGL Size Support for DSA3.0
  dmaengine: idxd: Expose DSA3.0 capabilities through sysfs
  dmaengine: sh: rz-dmac: Make channel irq local
  dmaengine: pl08x: Fix comment stating the difference between PL080 and PL081
  ...
</content>
</entry>
<entry>
<title>Documentation: use a source-read extension for the index link boilerplate</title>
<updated>2026-01-23T18:59:34Z</updated>
<author>
<name>Jani Nikula</name>
<email>jani.nikula@intel.com</email>
</author>
<published>2026-01-23T14:31:49Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=a592a36e49372172d7c7551ec19ed18184c935e1'/>
<id>urn:sha1:a592a36e49372172d7c7551ec19ed18184c935e1</id>
<content type='text'>
The root document usually has a special :ref:`genindex` link to the
generated index. This is also the case for Documentation/index.rst. The
other index.rst files deeper in the directory hierarchy usually don't.

For SPHINXDIRS builds, the root document isn't Documentation/index.rst,
but some other index.rst in the hierarchy. Currently they have a
".. only::" block to add the index link when doing SPHINXDIRS html
builds.

This is obviously very tedious and repetitive. The link is also added to
all index.rst files in the hierarchy for SPHINXDIRS builds, not just the
root document.

Put the boilerplate in a sphinx-includes/subproject-index.rst file, and
include it at the end of the root document for subproject builds in an
ad-hoc source-read extension defined in conf.py.

For now, keep having the boilerplate in translations, because this
approach currently doesn't cover translated index link headers.

Cc: Jonathan Corbet &lt;corbet@lwn.net&gt;
Cc: Mauro Carvalho Chehab &lt;mchehab@kernel.org&gt;
Cc: Randy Dunlap &lt;rdunlap@infradead.org&gt;
Signed-off-by: Jani Nikula &lt;jani.nikula@intel.com&gt;
Tested-by: Mauro Carvalho Chehab &lt;mchehab+huawei@kernel.org&gt;
Reviewed-by: Mauro Carvalho Chehab &lt;mchehab+huawei@kernel.org&gt;
[jc: did s/doctree/kern_doc_dir/ ]
Signed-off-by: Jonathan Corbet &lt;corbet@lwn.net&gt;
Message-ID: &lt;20260123143149.2024303-1-jani.nikula@intel.com&gt;
</content>
</entry>
<entry>
<title>docs: dmaengine: add explanation for phys field in dma_async_tx_descriptor structure</title>
<updated>2025-12-16T15:53:20Z</updated>
<author>
<name>Bhanu Seshu Kumar Valluri</name>
<email>bhanuseshukumar@gmail.com</email>
</author>
<published>2025-11-13T06:49:37Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=08be54a9e56f9523b50d1923a94a48ef5890c0bc'/>
<id>urn:sha1:08be54a9e56f9523b50d1923a94a48ef5890c0bc</id>
<content type='text'>
Describe the need to initialize the phys field in the dma_async_tx_descriptor
structure during its initialization.

Signed-off-by: Bhanu Seshu Kumar Valluri &lt;bhanuseshukumar@gmail.com&gt;
Reviewed-by: Bagas Sanjaya &lt;bagasdotme@gmail.com&gt;
Link: https://patch.msgid.link/20251113064937.8735-1-bhanuseshukumar@gmail.com
Signed-off-by: Vinod Koul &lt;vkoul@kernel.org&gt;
</content>
</entry>
<entry>
<title>Merge tag 'dmaengine-6.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine</title>
<updated>2025-06-05T15:49:30Z</updated>
<author>
<name>Linus Torvalds</name>
<email>torvalds@linux-foundation.org</email>
</author>
<published>2025-06-05T15:49:30Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=bfdf35c5dc6267f70f76abddfacface4dd3b9ac0'/>
<id>urn:sha1:bfdf35c5dc6267f70f76abddfacface4dd3b9ac0</id>
<content type='text'>
Pull dmaengine updates from Vinod Koul:
 "A fairly small update for the dmaengine subsystem. This has a new ARM
  dmaengine driver and couple of new device support and few driver
  changes:

  New support:
   - Renesas RZ/V2H(P) dma support for r9a09g057
   - Arm DMA-350 driver
   - Tegra Tegra264 ADMA support

  Updates:
   - AMD ptdma driver code removal and optimizations
   - Freescale edma error interrupt handler support"

* tag 'dmaengine-6.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/vkoul/dmaengine: (27 commits)
  dmaengine: idxd: Remove unused pointer and macro
  arm64: dts: renesas: r9a09g057: Add DMAC nodes
  dmaengine: sh: rz-dmac: Add RZ/V2H(P) support
  dmaengine: sh: rz-dmac: Allow for multiple DMACs
  irqchip/renesas-rzv2h: Add rzv2h_icu_register_dma_req()
  dt-bindings: dma: rz-dmac: Document RZ/V2H(P) family of SoCs
  dt-bindings: dma: rz-dmac: Restrict properties for RZ/A1H
  dmaengine: idxd: Narrow the restriction on BATCH to ver. 1 only
  dmaengine: ti: Add NULL check in udma_probe()
  fsldma: Set correct dma_mask based on hw capability
  dmaengine: idxd: Check availability of workqueue allocated by idxd wq driver before using
  dmaengine: xilinx_dma: Set dma_device directions
  dmaengine: tegra210-adma: Add Tegra264 support
  dt-bindings: Document Tegra264 ADMA support
  dmaengine: dw-edma: Add HDMA NATIVE map check
  dmaegnine: fsl-edma: add edma error interrupt handler
  dt-bindings: dma: fsl-edma: increase maxItems of interrupts and interrupt-names
  dmaengine: ARM_DMA350 should depend on ARM/ARM64
  dt-bindings: dma: qcom,bam: Document dma-coherent property
  dmaengine: Add Arm DMA-350 driver
  ...
</content>
</entry>
<entry>
<title>docs: dmaengine: add explanation for DMA_ASYNC_TX capability</title>
<updated>2025-05-07T07:02:57Z</updated>
<author>
<name>Kendra Moore</name>
<email>kendra.j.moore3443@gmail.com</email>
</author>
<published>2025-04-21T01:02:05Z</published>
<link rel='alternate' type='text/html' href='https://git.zx2c4.com/wireguard-linux/commit/?id=fea017a99ba5e28b23f8249c04b2f475e4ed1fb3'/>
<id>urn:sha1:fea017a99ba5e28b23f8249c04b2f475e4ed1fb3</id>
<content type='text'>
This patch replaces the TODO for DMA_ASYNC_TX in the DMA engine
provider documentation. The flag is automatically set by the DMA
framework when a device supports key asynchronous memory-to-memory
operations such as memcpy, memset, xor, pq, xor_val, and pq_val.

It must not be set by drivers directly.

Signed-off-by: Kendra Moore &lt;kendra.j.moore3443@gmail.com&gt;
Signed-off-by: Jonathan Corbet &lt;corbet@lwn.net&gt;
Message-ID: &lt;20250421010205.84719-1-kendra.j.moore3443@gmail.com&gt;
</content>
</entry>
</feed>
