<feed xmlns='http://www.w3.org/2005/Atom'>
<title>linux-toradex.git/include/linux/mlx5, branch master</title>
<subtitle>Linux kernel for Apalis and Colibri modules</subtitle>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/'/>
<entry>
<title>net/mlx5e: Serialize TC and IPsec offload exclusion counters</title>
<updated>2026-11-10T23:33:09+00:00</updated>
<author>
<name>Cosmin Ratiu</name>
<email>cratiu@nvidia.com</email>
</author>
<published>2026-11-05T21:16:36+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=717dec565a5e7feab544da5aa77db9da71bc93fe'/>
<id>717dec565a5e7feab544da5aa77db9da71bc93fe</id>
<content type='text'>
The counters enforcing TC and IPsec packet offload mutual exclusion are
not consistently serialized. IPsec admission conditionally takes
mode_lock for writing, while TC admission holds it only for read.
Neither release path takes the lock, and devices without an eswitch
cannot rely on it at all.

These are per-core-device counters, separate from the eswitch mode and
encapsulation reservations protected by mode_lock. Move them into
mdev-&gt;offload_block and protect their checks, increments and decrements
with a dedicated spinlock that is available even without an eswitch.
A spinlock is used because some of these functions are called from
atomic context and is a good fit because the critical sections are and
will stay really short.

IPsec admission then no longer needs mode_lock to update these counters.
Add warnings for unbalanced releases. Keep mlx5_esw_lock() for eswitch
operation admission and temporary mode/encapsulation reservations.

This was found by upstream Sashiko during review of [1], and spun-off
into a fix from the larger patch.

[1]
https://lore.kernel.org/netdev/178907726586.219967.4695695146458742597@kernel.org/

Fixes: c8e350e62fc5 ("net/mlx5e: Make TC and IPsec offloads mutually exclusive on a netdev")
Signed-off-by: Cosmin Ratiu &lt;cratiu@nvidia.com&gt;
Reviewed-by: Jianbo Liu &lt;jianbol@nvidia.com&gt;
Reviewed-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20261105211637.1083978-9-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski &lt;kuba@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The counters enforcing TC and IPsec packet offload mutual exclusion are
not consistently serialized. IPsec admission conditionally takes
mode_lock for writing, while TC admission holds it only for read.
Neither release path takes the lock, and devices without an eswitch
cannot rely on it at all.

These are per-core-device counters, separate from the eswitch mode and
encapsulation reservations protected by mode_lock. Move them into
mdev-&gt;offload_block and protect their checks, increments and decrements
with a dedicated spinlock that is available even without an eswitch.
A spinlock is used because some of these functions are called from
atomic context and is a good fit because the critical sections are and
will stay really short.

IPsec admission then no longer needs mode_lock to update these counters.
Add warnings for unbalanced releases. Keep mlx5_esw_lock() for eswitch
operation admission and temporary mode/encapsulation reservations.

This was found by upstream Sashiko during review of [1], and spun-off
into a fix from the larger patch.

[1]
https://lore.kernel.org/netdev/178907726586.219967.4695695146458742597@kernel.org/

Fixes: c8e350e62fc5 ("net/mlx5e: Make TC and IPsec offloads mutually exclusive on a netdev")
Signed-off-by: Cosmin Ratiu &lt;cratiu@nvidia.com&gt;
Reviewed-by: Jianbo Liu &lt;jianbol@nvidia.com&gt;
Reviewed-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20261105211637.1083978-9-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski &lt;kuba@kernel.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>Merge tag 'vfio-v7.4-rc1' of https://github.com/awilliam/linux-vfio</title>
<updated>2026-10-24T00:22:52+00:00</updated>
<author>
<name>Linus Torvalds</name>
<email>torvalds@linux-foundation.org</email>
</author>
<published>2026-10-24T00:22:52+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=26344c0dd32af8790d8474d233619db825441138'/>
<id>26344c0dd32af8790d8474d233619db825441138</id>
<content type='text'>
Pull VFIO updates from Alex Williamson:

 - Fix missing selftest &lt;libgen.h&gt; include for basename() to support
   non-glibc builds (Hisam Mehboob)

 - Add vfio selftest documentation covering the libvfio helper library
   and guidance for developers to aid in adding and understanding the
   framework (David Matlack)

 - Fix vfio cdev open path to prevent a second open from clobbering the
   vf_token or clearing the KVM reference. Add a selftest to avoid
   future recurrences (Alex Williamson)

 - Extend vfio selftest mix_and_match timeouts, in particular to account
   for systems where generated faults can induce heavy SMM load,
   delaying internal timeouts. Also extend timeout for execution via
   runner to account for various tests exceeding default 45s timeout
   (Alex Williamson)

 - Add tracking for non-reserved mappings into a vfio_dma object in
   order to optimize the tear-down paths where per-page tests can be
   entirely skipped. This resolves a soft lockup for large memory GPUs
   (Samuel Crossley)

 - Add vfio selftest driver for mlx5, providing another target for
   exercising DMA and interrupts against a physical device (Jason
   Gunthorpe)

 - Hardening against read-only region promotion, preserving vm_pgoff for
   unmap_mapping_range(), and reject non-shared MMIO mmaps as necessary
   through vfio-platform, vfio-fsl-mc, and vfio-cdx. These prevent
   future issues, though none are currently reachable (Abdifatah Suruur)

 - Fix 8-byte ioeventfd support, removing a static error test that was
   erroneously left in place when generalizing 8-byte io accesses (Li
   RongQing)

 - Fix missed RCU drain on module exit (Jiale Yao)

 - Fix off-by-one boundary check in migration support within the mlx5
   vfio-pci variant driver (Seongjun Hong)

 - Remove drvdata usage convention across vfio-pci-core, allowing
   variant drivers to freely create their own conventions and enabling a
   future Rust-based variant driver. A DRM topic branch is also merged
   in support of this effort, modifying the vga arbiter API to more
   cleanly remove the drvdata dependency (Simon Song)

 - Fix hisi_acc vfio-pci variant driver to avoid NULL pointer accesses
   when bound to unsupported devices and also disabling migration
   support when running on 64K host page size due to BAR mapping
   constraints (Longfang Liu)

 - Add device IDs for Crescent Island and Nova Lake devices to the Xe
   vfio-pci variant driver (Satyanarayana K V P)

 - Fix fallback mode in nvgrace-gpu variant driver to use symmetric open
   and close handlers to avoid reported use-after-free issue. Remove the
   resulting unnecessary test in the now single-purpose open function
   (Andrea Parri)

* tag 'vfio-v7.4-rc1' of https://github.com/awilliam/linux-vfio: (35 commits)
  vfio/nvgrace-gpu: Drop the usemem length check in open_device
  vfio/nvgrace-gpu: Use a core-only open_device in fallback mode
  vfio/xe: Add CRI and NVL device IDs
  hisi_acc_vfio_pci: reject live migration on 64KB page with QM_HW_V3 hardware
  hisi_acc_vfio_pci: fix NULL dereference in reset_prepare on PF passthrough
  vfio/pci: Remove the core dependency on driver data
  vfio/mlx5: Fix boundary check in set_report_output()
  vfio/pci: Drain eventfd RCU callbacks on module exit
  PCI/VGA: pass a data pointer to be used in the decode callback (v2)
  vfio/pci: Restore 8-byte ioeventfd support
  vfio/cdx: reject non-shared MMIO mmaps
  vfio/cdx: keep logical vm_pgoff in MMIO region mmap
  vfio/fsl-mc: keep logical vm_pgoff in MMIO region mmap
  vfio/platform: keep logical vm_pgoff in MMIO region mmap
  vfio/cdx: prevent read-only region mappings from becoming writable
  vfio/fsl-mc: prevent read-only region mappings from becoming writable
  vfio/platform: prevent read-only region mappings from becoming writable
  vfio: selftests: Verify a failed second open preserves the vf_token
  vfio: selftests: mlx5 driver - add send_msi support
  vfio: selftests: Add mlx5 driver - data path and memcpy ops
  ...
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Pull VFIO updates from Alex Williamson:

 - Fix missing selftest &lt;libgen.h&gt; include for basename() to support
   non-glibc builds (Hisam Mehboob)

 - Add vfio selftest documentation covering the libvfio helper library
   and guidance for developers to aid in adding and understanding the
   framework (David Matlack)

 - Fix vfio cdev open path to prevent a second open from clobbering the
   vf_token or clearing the KVM reference. Add a selftest to avoid
   future recurrences (Alex Williamson)

 - Extend vfio selftest mix_and_match timeouts, in particular to account
   for systems where generated faults can induce heavy SMM load,
   delaying internal timeouts. Also extend timeout for execution via
   runner to account for various tests exceeding default 45s timeout
   (Alex Williamson)

 - Add tracking for non-reserved mappings into a vfio_dma object in
   order to optimize the tear-down paths where per-page tests can be
   entirely skipped. This resolves a soft lockup for large memory GPUs
   (Samuel Crossley)

 - Add vfio selftest driver for mlx5, providing another target for
   exercising DMA and interrupts against a physical device (Jason
   Gunthorpe)

 - Hardening against read-only region promotion, preserving vm_pgoff for
   unmap_mapping_range(), and reject non-shared MMIO mmaps as necessary
   through vfio-platform, vfio-fsl-mc, and vfio-cdx. These prevent
   future issues, though none are currently reachable (Abdifatah Suruur)

 - Fix 8-byte ioeventfd support, removing a static error test that was
   erroneously left in place when generalizing 8-byte io accesses (Li
   RongQing)

 - Fix missed RCU drain on module exit (Jiale Yao)

 - Fix off-by-one boundary check in migration support within the mlx5
   vfio-pci variant driver (Seongjun Hong)

 - Remove drvdata usage convention across vfio-pci-core, allowing
   variant drivers to freely create their own conventions and enabling a
   future Rust-based variant driver. A DRM topic branch is also merged
   in support of this effort, modifying the vga arbiter API to more
   cleanly remove the drvdata dependency (Simon Song)

 - Fix hisi_acc vfio-pci variant driver to avoid NULL pointer accesses
   when bound to unsupported devices and also disabling migration
   support when running on 64K host page size due to BAR mapping
   constraints (Longfang Liu)

 - Add device IDs for Crescent Island and Nova Lake devices to the Xe
   vfio-pci variant driver (Satyanarayana K V P)

 - Fix fallback mode in nvgrace-gpu variant driver to use symmetric open
   and close handlers to avoid reported use-after-free issue. Remove the
   resulting unnecessary test in the now single-purpose open function
   (Andrea Parri)

* tag 'vfio-v7.4-rc1' of https://github.com/awilliam/linux-vfio: (35 commits)
  vfio/nvgrace-gpu: Drop the usemem length check in open_device
  vfio/nvgrace-gpu: Use a core-only open_device in fallback mode
  vfio/xe: Add CRI and NVL device IDs
  hisi_acc_vfio_pci: reject live migration on 64KB page with QM_HW_V3 hardware
  hisi_acc_vfio_pci: fix NULL dereference in reset_prepare on PF passthrough
  vfio/pci: Remove the core dependency on driver data
  vfio/mlx5: Fix boundary check in set_report_output()
  vfio/pci: Drain eventfd RCU callbacks on module exit
  PCI/VGA: pass a data pointer to be used in the decode callback (v2)
  vfio/pci: Restore 8-byte ioeventfd support
  vfio/cdx: reject non-shared MMIO mmaps
  vfio/cdx: keep logical vm_pgoff in MMIO region mmap
  vfio/fsl-mc: keep logical vm_pgoff in MMIO region mmap
  vfio/platform: keep logical vm_pgoff in MMIO region mmap
  vfio/cdx: prevent read-only region mappings from becoming writable
  vfio/fsl-mc: prevent read-only region mappings from becoming writable
  vfio/platform: prevent read-only region mappings from becoming writable
  vfio: selftests: Verify a failed second open preserves the vf_token
  vfio: selftests: mlx5 driver - add send_msi support
  vfio: selftests: Add mlx5 driver - data path and memcpy ops
  ...
</pre>
</div>
</content>
</entry>
<entry>
<title>Merge branch 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux</title>
<updated>2026-10-16T01:48:40+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>kuba@kernel.org</email>
</author>
<published>2026-10-16T01:48:39+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=9214626223a04a992746ee81c98347d6b739a309'/>
<id>9214626223a04a992746ee81c98347d6b739a309</id>
<content type='text'>
Tariq Toukan says:

====================
mlx5-next updates 2026-10-11

* 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux:
  net/mlx5: IFC, Add capability bits for virtual LAG
  net/mlx5: Add cap_tx_speed and effective_tx_speed support to IFC
====================

Link: https://patch.msgid.link/20261011113534.802524-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski &lt;kuba@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Tariq Toukan says:

====================
mlx5-next updates 2026-10-11

* 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux:
  net/mlx5: IFC, Add capability bits for virtual LAG
  net/mlx5: Add cap_tx_speed and effective_tx_speed support to IFC
====================

Link: https://patch.msgid.link/20261011113534.802524-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski &lt;kuba@kernel.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/mlx5: IFC, Add capability bits for virtual LAG</title>
<updated>2026-10-11T11:06:18+00:00</updated>
<author>
<name>Shay Drory</name>
<email>shayd@nvidia.com</email>
</author>
<published>2026-10-05T08:40:47+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=41de84543574f2923776790f608c56389aed972e'/>
<id>41de84543574f2923776790f608c56389aed972e</id>
<content type='text'>
Virtual LAG can also be built over virtual uplinks rather than physical
ports.
The support for this virtual LAG is indicated by multi_uplink_eswitch
and port_select_eswitch caps bit.

Signed-off-by: Shay Drory &lt;shayd@nvidia.com&gt;
Reviewed-by: Mark Bloch &lt;mbloch@nvidia.com&gt;
Reviewed-by: Moshe Shemesh &lt;moshe@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20261005084047.237081-3-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Virtual LAG can also be built over virtual uplinks rather than physical
ports.
The support for this virtual LAG is indicated by multi_uplink_eswitch
and port_select_eswitch caps bit.

Signed-off-by: Shay Drory &lt;shayd@nvidia.com&gt;
Reviewed-by: Mark Bloch &lt;mbloch@nvidia.com&gt;
Reviewed-by: Moshe Shemesh &lt;moshe@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20261005084047.237081-3-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/mlx5: Add cap_tx_speed and effective_tx_speed support to IFC</title>
<updated>2026-10-11T11:06:18+00:00</updated>
<author>
<name>Or Har-Toov</name>
<email>ohartoov@nvidia.com</email>
</author>
<published>2026-10-05T08:40:46+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=4c5679222f6b9bb6800d32eb8256b43d164c3e31'/>
<id>4c5679222f6b9bb6800d32eb8256b43d164c3e31</id>
<content type='text'>
Introduce the cap_tx_speed and effective_tx_speed fields to the query
and modify_vport_state structures.

Add the esw_vport_state_cap_tx_speed and
esw_vport_state_effective_tx_speed capability bits, indicating the
firmware supports modifying the cap_tx_speed and effective_tx_speed
fields, via the MODIFY_VPORT_STATE command.

Signed-off-by: Or Har-Toov &lt;ohartoov@nvidia.com&gt;
Reviewed-by: Mark Bloch &lt;mbloch@nvidia.com&gt;
Reviewed-by: Nimrod Oren &lt;noren@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20261005084047.237081-2-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Introduce the cap_tx_speed and effective_tx_speed fields to the query
and modify_vport_state structures.

Add the esw_vport_state_cap_tx_speed and
esw_vport_state_effective_tx_speed capability bits, indicating the
firmware supports modifying the cap_tx_speed and effective_tx_speed
fields, via the MODIFY_VPORT_STATE command.

Signed-off-by: Or Har-Toov &lt;ohartoov@nvidia.com&gt;
Reviewed-by: Mark Bloch &lt;mbloch@nvidia.com&gt;
Reviewed-by: Nimrod Oren &lt;noren@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20261005084047.237081-2-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>Merge branch 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux</title>
<updated>2026-10-07T01:50:37+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>kuba@kernel.org</email>
</author>
<published>2026-10-07T01:50:36+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=8e37d7bec7968f4b63b47e5b478c3d5b98f02793'/>
<id>8e37d7bec7968f4b63b47e5b478c3d5b98f02793</id>
<content type='text'>
Tariq Toukan says:

====================
mlx5: Move data direct infrastructure to mlx5_core

Data direct support currently lives in mlx5_ib, which owns the data
direct PCI driver, device matching and resource management. To prepare
for data direct support in netdev, move this infrastructure to mlx5_core
and provide a common interface for its consumers.

The series first replaces the direct IB unbind callback with a blocking
notifier chain, allowing multiple consumers to receive unbind events.
It then separates consumer registration from initialization, consolidates
the per-device state and resources, and keeps IB-specific locking and MR
teardown in mlx5_ib.

With these preparations in place, the final patch moves the shared
implementation and PCI driver to mlx5_core. Data direct initialization
and resource creation run during core load, before consumers attach;
cleanup runs during core unload, after consumers detach.

The data direct PCI driver is also explicitly named "mlx5_dd", replacing
the previous "mlx5_ib" name derived from KBUILD_MODNAME.

This series prepares the shared infrastructure for netdev support.
The netdev consumer will follow in a separate series.

* 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux:
  mlx5: Move data direct implementation to mlx5_core
  RDMA/mlx5: Pull data_direct resource creation in init phase
  RDMA/mlx5: Consolidate data direct state into one object
  RDMA/mlx5: Extract IB specific lock out of data direct
  RDMA/mlx5: Add new registration stage for data direct users
  RDMA/mlx5: Move and rename data direct resource functions
  RDMA/mlx5: Give the data direct PCI driver a proper name
  RDMA/mlx5: Make the data direct VUID query self-contained
  RDMA/mlx5: Add mlx5_data_direct_supported() helper
  RDMA/mlx5: Notify data direct users via a notifier chain on unbind
====================

Link: https://patch.msgid.link/20261004064921.213888-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski &lt;kuba@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Tariq Toukan says:

====================
mlx5: Move data direct infrastructure to mlx5_core

Data direct support currently lives in mlx5_ib, which owns the data
direct PCI driver, device matching and resource management. To prepare
for data direct support in netdev, move this infrastructure to mlx5_core
and provide a common interface for its consumers.

The series first replaces the direct IB unbind callback with a blocking
notifier chain, allowing multiple consumers to receive unbind events.
It then separates consumer registration from initialization, consolidates
the per-device state and resources, and keeps IB-specific locking and MR
teardown in mlx5_ib.

With these preparations in place, the final patch moves the shared
implementation and PCI driver to mlx5_core. Data direct initialization
and resource creation run during core load, before consumers attach;
cleanup runs during core unload, after consumers detach.

The data direct PCI driver is also explicitly named "mlx5_dd", replacing
the previous "mlx5_ib" name derived from KBUILD_MODNAME.

This series prepares the shared infrastructure for netdev support.
The netdev consumer will follow in a separate series.

* 'mlx5-next' of git://git.kernel.org/pub/scm/linux/kernel/git/mellanox/linux:
  mlx5: Move data direct implementation to mlx5_core
  RDMA/mlx5: Pull data_direct resource creation in init phase
  RDMA/mlx5: Consolidate data direct state into one object
  RDMA/mlx5: Extract IB specific lock out of data direct
  RDMA/mlx5: Add new registration stage for data direct users
  RDMA/mlx5: Move and rename data direct resource functions
  RDMA/mlx5: Give the data direct PCI driver a proper name
  RDMA/mlx5: Make the data direct VUID query self-contained
  RDMA/mlx5: Add mlx5_data_direct_supported() helper
  RDMA/mlx5: Notify data direct users via a notifier chain on unbind
====================

Link: https://patch.msgid.link/20261004064921.213888-1-tariqt@nvidia.com
Signed-off-by: Jakub Kicinski &lt;kuba@kernel.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>mlx5: Move data direct implementation to mlx5_core</title>
<updated>2026-10-01T15:40:42+00:00</updated>
<author>
<name>Dragos Tatulea</name>
<email>dtatulea@nvidia.com</email>
</author>
<published>2026-09-30T11:42:27+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=00290b9ff59ef47b990d26a6b1a8b2ea6dedf5e9'/>
<id>00290b9ff59ef47b990d26a6b1a8b2ea6dedf5e9</id>
<content type='text'>
As data direct will also be used by netdev devices, move the data direct
code to mlx5_core. Now mlx5_core is responsible for data direct
initialization and resource creation.

Now data_direct is initialized in mlx5_load() before attaching consumers
and clean it up in mlx5_unload() after detaching them. The data direct
PCI driver is moved and registered in mlx5_core.

IB specific work stays in mlx5_ib: notifier registration, mutex
initialization and MR list initialization.

Failure to query the VUID is now fatal because this is done only on
devices that support data direct. Other than this, this patch has no
additional functional changes.

Signed-off-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Link: https://patch.msgid.link/20260930114227.139274-11-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
As data direct will also be used by netdev devices, move the data direct
code to mlx5_core. Now mlx5_core is responsible for data direct
initialization and resource creation.

Now data_direct is initialized in mlx5_load() before attaching consumers
and clean it up in mlx5_unload() after detaching them. The data direct
PCI driver is moved and registered in mlx5_core.

IB specific work stays in mlx5_ib: notifier registration, mutex
initialization and MR list initialization.

Failure to query the VUID is now fatal because this is done only on
devices that support data direct. Other than this, this patch has no
additional functional changes.

Signed-off-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Link: https://patch.msgid.link/20260930114227.139274-11-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>RDMA/mlx5: Consolidate data direct state into one object</title>
<updated>2026-10-01T15:40:42+00:00</updated>
<author>
<name>Dragos Tatulea</name>
<email>dtatulea@nvidia.com</email>
</author>
<published>2026-09-30T11:42:25+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=f6c452b311f0e4a8fc2a0c287dde298d5492112a'/>
<id>f6c452b311f0e4a8fc2a0c287dde298d5492112a</id>
<content type='text'>
There are three per device data direct objects: mlx5_data_direct_dev,
mlx5_data_direct_resources and mlx5_data_direct_registration with
separate allocations. This patch consolidates all of them into the
registration object.

For this, a few things needed to happen:

- mlx5_ib_dev keeps a pointer to mlx5_data_direct.
- mlx5_data_direct_dev is now accessible from mlx5_data_direct.
  mlx5_data_direct_get_dev() is added as a helper to access
  the device pointer.
- Since mlx5_data_direct is part of mlx5_data_direct_registration,
  getting to the registration is a simple container_of. No need
  for searching the list.

Since mlx5_data_direct_resources no longer has an accurate name,
this patch renames it.

This is a preparation for the upcoming move of data_direct to
mlx5_core.

This patch has no functional changes.

Signed-off-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20260930114227.139274-9-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
There are three per device data direct objects: mlx5_data_direct_dev,
mlx5_data_direct_resources and mlx5_data_direct_registration with
separate allocations. This patch consolidates all of them into the
registration object.

For this, a few things needed to happen:

- mlx5_ib_dev keeps a pointer to mlx5_data_direct.
- mlx5_data_direct_dev is now accessible from mlx5_data_direct.
  mlx5_data_direct_get_dev() is added as a helper to access
  the device pointer.
- Since mlx5_data_direct is part of mlx5_data_direct_registration,
  getting to the registration is a simple container_of. No need
  for searching the list.

Since mlx5_data_direct_resources no longer has an accurate name,
this patch renames it.

This is a preparation for the upcoming move of data_direct to
mlx5_core.

This patch has no functional changes.

Signed-off-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20260930114227.139274-9-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>RDMA/mlx5: Add new registration stage for data direct users</title>
<updated>2026-10-01T15:40:42+00:00</updated>
<author>
<name>Dragos Tatulea</name>
<email>dtatulea@nvidia.com</email>
</author>
<published>2026-09-30T11:42:23+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=1c879b1e9f2ed620523e17ca239683bc8029e143'/>
<id>1c879b1e9f2ed620523e17ca239683bc8029e143</id>
<content type='text'>
For the upcoming move of data direct to mlx5_core driver,
the API needs to be split into init and register stages.
That's because only the core will do the init phase but
both IB and netdev devices will need to register.

While data direct still lives in mlx5_ib, do all steps in
mlx5_ib_data_direct_init().

While at it, make mlx5_data_direct_query_vuid() private
as it is no longer needed.

This patch has no functional changes. It is a preparation
for moving data direct infrastructure to mlx5_core. For
this reason the new functions are added to the new header.

Signed-off-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Reviewed-by: Leon Romanovsky &lt;leonro@nvidia.com&gt;
Reviewed-by: Cosmin Ratiu &lt;cratiu@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20260930114227.139274-7-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
For the upcoming move of data direct to mlx5_core driver,
the API needs to be split into init and register stages.
That's because only the core will do the init phase but
both IB and netdev devices will need to register.

While data direct still lives in mlx5_ib, do all steps in
mlx5_ib_data_direct_init().

While at it, make mlx5_data_direct_query_vuid() private
as it is no longer needed.

This patch has no functional changes. It is a preparation
for moving data direct infrastructure to mlx5_core. For
this reason the new functions are added to the new header.

Signed-off-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Reviewed-by: Leon Romanovsky &lt;leonro@nvidia.com&gt;
Reviewed-by: Cosmin Ratiu &lt;cratiu@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20260930114227.139274-7-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>RDMA/mlx5: Make the data direct VUID query self-contained</title>
<updated>2026-10-01T15:40:42+00:00</updated>
<author>
<name>Dragos Tatulea</name>
<email>dtatulea@nvidia.com</email>
</author>
<published>2026-09-30T11:42:20+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=ebbb9f87689998157e7310e96f585bc293f6fcb0'/>
<id>ebbb9f87689998157e7310e96f585bc293f6fcb0</id>
<content type='text'>
mlx5_cmd_query_vuid() in cmd.c is generic over the data_direct flag but
its only caller is the data direct init path, which always passes
data_direct=true.

Move it into data_direct.c as mlx5_data_direct_query_vuid(), dropping
the always-true parameter and hardcoding data_direct=1. This groups the
query with the rest of the data direct code so that it travels together
when the data direct matching code is relocated to mlx5_core, and lets
it become static there once the query is folded into the core bind path.

This function will become private in a subsequent patch.

No functional change.

Signed-off-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Reviewed-by: Leon Romanovsky &lt;leonro@nvidia.com&gt;
Reviewed-by: Cosmin Ratiu &lt;cratiu@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20260930114227.139274-4-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
mlx5_cmd_query_vuid() in cmd.c is generic over the data_direct flag but
its only caller is the data direct init path, which always passes
data_direct=true.

Move it into data_direct.c as mlx5_data_direct_query_vuid(), dropping
the always-true parameter and hardcoding data_direct=1. This groups the
query with the rest of the data direct code so that it travels together
when the data direct matching code is relocated to mlx5_core, and lets
it become static there once the query is folded into the core bind path.

This function will become private in a subsequent patch.

No functional change.

Signed-off-by: Dragos Tatulea &lt;dtatulea@nvidia.com&gt;
Reviewed-by: Leon Romanovsky &lt;leonro@nvidia.com&gt;
Reviewed-by: Cosmin Ratiu &lt;cratiu@nvidia.com&gt;
Signed-off-by: Tariq Toukan &lt;tariqt@nvidia.com&gt;
Link: https://patch.msgid.link/20260930114227.139274-4-tariqt@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leon@kernel.org&gt;
</pre>
</div>
</content>
</entry>
</feed>
