Commit graph

1,444,037 commits

Author SHA1 Message Date
Greg Kroah-Hartman
2f0dad8baa Merge branch 'staging-linus' into branch 'staging-next'
There were merge conflicts in:
	drivers/staging/rtl8723bs/core/rtw_ieee80211.c
	drivers/staging/sm750fb/sm750_hw.c

that were causing problems, so fix this up in here so that going forward
everyone is on the same page and linux-next does not have problems.

Reported-by: Mark Brown <broonie@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:15:08 -08:00
Hardik Phalet
ba30a9f9be staging: greybus: audio: drop stale TODO comment
Modules are removed from modules_list in gb_audio_manager_remove() and
gb_audio_manager_remove_all() before kobject_put(). The TODO suggesting
list deletion in the kobject release callback is stale and misleading.

Signed-off-by: Hardik Phalet <hardik.phalet@pm.me>
Link: https://patch.msgid.link/20260224185421.824210-3-hardik.phalet@pm.me
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:05:07 -08:00
Hardik Phalet
785fe65012 staging: greybus: audio: remove unused gb_audio_manager_get_module()
gb_audio_manager_get_module() has no in-tree callers so remove the
unused function to avoid carrying dead code.

Signed-off-by: Hardik Phalet <hardik.phalet@pm.me>
Link: https://patch.msgid.link/20260224185421.824210-2-hardik.phalet@pm.me
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:04:51 -08:00
Artem Lytkin
0886fb23a8 staging: most: dim2: remove unnecessary string indirection in dev_err
Replace dev_err(&pdev->dev, "%s\n", "clk_prepare_enable failed") with
the direct format string dev_err(&pdev->dev, "clk_prepare_enable
failed\n"). The extra level of indirection through %s is unnecessary.

Signed-off-by: Artem Lytkin <iprintercanon@gmail.com>
Link: https://patch.msgid.link/20260224180750.28468-4-iprintercanon@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:03:40 -08:00
Artem Lytkin
b7a013c125 staging: most: dim2: use dev_err_probe and proper error codes for clock
Replace hardcoded -EFAULT returns with dev_err_probe() and PTR_ERR()
when devm_clk_get() fails in fsl_mx6_enable(). This ensures the
correct error code is propagated (e.g. -EPROBE_DEFER for deferred
probing) and avoids log noise during probe deferral.

Signed-off-by: Artem Lytkin <iprintercanon@gmail.com>
Link: https://patch.msgid.link/20260224180750.28468-3-iprintercanon@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:03:40 -08:00
Artem Lytkin
24b28dc218 staging: most: dim2: replace IS_ERR_OR_NULL with IS_ERR for devm_clk_get
devm_clk_get() never returns NULL, so IS_ERR_OR_NULL() checks are
unnecessary. Replace them with IS_ERR() for both the "mlb" and
"pll8_mlb" clock lookups in fsl_mx6_enable().

Signed-off-by: Artem Lytkin <iprintercanon@gmail.com>
Link: https://patch.msgid.link/20260224180750.28468-2-iprintercanon@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:03:40 -08:00
Dhyan K Prajapati
4895acd1b4 staging: rtl8723bs: fix function header alignment
fixed header alignment in hal_sdio_get_cmd_addr_8723b that violated
kernel coding style, cleaned the header by pulling arguments onto the
same line.

Signed-off-by: Dhyan K Prajapati <dhyan19022009@gmail.com>
Link: https://patch.msgid.link/20260225085747.9639-1-dhyan19022009@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:03:00 -08:00
Shubham Chakraborty
14c91651df staging: greybus: audio: Use sysfs_emit in show functions
Refactor sprintf to sysfs_emit in all show functions of the greybus
audio manager module. This follows the standard kernel practice of
using sysfs_emit for sysfs attributes.

Signed-off-by: Shubham Chakraborty <chakrabortyshubham66@gmail.com>
Link: https://patch.msgid.link/20260225071212.9050-1-chakrabortyshubham66@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:02:26 -08:00
Khasar Munkh-Erdene
393b267e6f staging: rtl8723bs: fix spacing around operators in rtw_recv.c
Fix coding style issues by adding missing spaces around binary and
ternary operators, as reported by checkpatch.pl.

Signed-off-by: Khasar Munkh-Erdene <02khasar@gmail.com>
Link: https://patch.msgid.link/20260225032909.72799-1-02khasar@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 07:02:20 -08:00
Jens Glathe
b0829cd6a6 arm64: dts: qcom: x1-vivobook-s15: add Purwa-compatible device tree
The ASUS Vivobook S15 (S5507) [1] is available with Hamoa and Purwa SoC.

Add the Purwa-based device tree.

[1]: https://www.asus.com/de/laptops/for-home/vivobook/asus-vivobook-s-15-s5507/techspec/

Tested-by: Colin K. Williams <colin@li-nk.org>
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Signed-off-by: Jens Glathe <jens.glathe@oldschoolsolutions.biz>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Link: https://lore.kernel.org/r/20260214-b4-vivobook-v3-4-3c88065bbf77@oldschoolsolutions.biz
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2026-02-25 09:00:55 -06:00
Jens Glathe
da35a8931f arm64: dts: qcom: x1-vivobook-s15: create a common dtsi for Hamoa and Purwa variants
The Asus VivoBook S15 S5507QA is sold with x1e-78-100 and x1p-42-100 SKUs.
Put the common part into one dtsi file to be included for model-specific dts.
Include the common part in the existing Vivobook S15 device tree.

Tested-by: Colin K. Williams <colin@li-nk.org>
Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Signed-off-by: Jens Glathe <jens.glathe@oldschoolsolutions.biz>
Tested-by: Maud Spierings <maud_spierings@hotmail.com>
Link: https://lore.kernel.org/r/20260214-b4-vivobook-v3-3-3c88065bbf77@oldschoolsolutions.biz
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2026-02-25 09:00:23 -06:00
Jens Glathe
1fdbcc6d3f dt-bindings: arm: qcom: Add ASUS Vivobook X1P42100 variant
The ASUS Vivobook S15 (S5507) [1] is available with Hamoa and Purwa SoC.
Add the Purwa-based variant:

asus,vivobook-s15-x1p4 compatible to Purwa SoC

[1]: https://www.asus.com/de/laptops/for-home/vivobook/asus-vivobook-s-15-s5507/techspec/

Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com>
Signed-off-by: Jens Glathe <jens.glathe@oldschoolsolutions.biz>
Link: https://lore.kernel.org/r/20260214-b4-vivobook-v3-1-3c88065bbf77@oldschoolsolutions.biz
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2026-02-25 09:00:23 -06:00
Jens Glathe
26b8661065 firmware: qcom: scm: allow QSEECOM on ASUS Vivobook X1P42100 variant
Enables access to EFI variables on this machine.

Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Tested-by: Colin K. Williams <colin@li-nk.org>
Signed-off-by: Jens Glathe <jens.glathe@oldschoolsolutions.biz>
Link: https://lore.kernel.org/r/20260214-b4-vivobook-v3-2-3c88065bbf77@oldschoolsolutions.biz
Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2026-02-25 09:00:10 -06:00
Nikolay Kulikov
9bf1804229 staging: rtl8723bs: remove recurring counter increment
The code:
	cnt += in_ie[cnt + 1] + 2;   /* get next */
is in both the "if" and "else" branches.

Remove this repetition, which will simplify the code and
improve readability.

Signed-off-by: Nikolay Kulikov <nikolayof23@gmail.com>
Link: https://patch.msgid.link/20260224164445.18316-1-nikolayof23@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 06:59:12 -08:00
Gabriel Windlin
0c038cb19a staging: rtl8723bs: remove multiple blank lines
Remove multiple consecutive blank lines to adhere to the Linux kernel
coding style. Issues reported by checkpatch.pl.

Signed-off-by: Gabriel Windlin <gawindlin@gmail.com>
Link: https://patch.msgid.link/20260224161319.89187-1-gawindlin@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2026-02-25 06:59:12 -08:00
Panagiotis Foliadis
e9fb2028f1 ALSA: hda/intel: increase default bdl_pos_adj for Nvidia controllers
The default bdl_pos_adj of 32 for Nvidia HDA controllers is
insufficient on GA102 (and likely other recent Nvidia GPUs) after S3
suspend/resume. The controller's DMA timing degrades after resume,
causing premature IRQ detection in azx_position_ok() which results in
silent HDMI/DP audio output despite userspace reporting a valid
playback state and correct ELD data.

Increase bdl_pos_adj to 64 for AZX_DRIVER_NVIDIA, matching the value
already used by Intel Apollo Lake for the same class of timing issue.

Cc: stable@vger.kernel.org
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=221069
Suggested-by: Charalampos Mitrodimas <charmitro@posteo.net>
Signed-off-by: Panagiotis Foliadis <pfoliadis@posteo.net>
Link: https://patch.msgid.link/20260225-nvidia-audio-fix-v1-1-b1383c37ec49@posteo.net
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2026-02-25 15:56:57 +01:00
Nathan Chancellor
ab6088e7a9 lib/Kconfig.debug: Require a release version of LLVM 22 for context analysis
Using a prerelease version as a minimum supported version for
CONFIG_WARN_CONTEXT_ANALYSIS was reasonable to do while LLVM 22 was the
development version so that people could immediately build from main and
start testing and validating this in their own code. However, it can be
problematic when using prerelease versions of LLVM 22, such as Android
clang 22.0.1 (the current android mainline compiler) or when bisecting
LLVM between llvmorg-22-init and llvmorg-23-init, to build the kernel,
as all compiler fixes for the context analysis may not be present,
potentially resulting in warnings that can easily turn into errors.

Now that LLVM 22 is released as 22.1.0, upgrade the check to require at
least this version to ensure that a user's toolchain actually has all
the changes needed for a smooth experience with context analysis.

Fixes: 3269701cb2 ("compiler-context-analysis: Add infrastructure for Context Analysis with Clang")
Signed-off-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Marco Elver <elver@google.com>
Link: https://patch.msgid.link/20260224-bump-clang-ver-context-analysis-v1-1-16cc7a90a040@kernel.org
2026-02-25 15:36:02 +01:00
Tommaso Cucinotta
2e7af19269 sched/deadline: Add reporting of runtime left & abs deadline to sched_getattr() for DEADLINE tasks
The SCHED_DEADLINE scheduler allows reading the statically configured
run-time, deadline, and period parameters through the sched_getattr()
system call. However, there is no immediate way to access, from user space,
the current parameters used within the scheduler: the instantaneous runtime
left in the current cycle, as well as the current absolute deadline.

The `flags' sched_getattr() parameter, so far mandated to contain zero,
now supports the SCHED_GETATTR_FLAG_DL_DYNAMIC=1 flag, to request
retrieval of the leftover runtime and absolute deadline, converted to a
CLOCK_MONOTONIC reference, instead of the statically configured parameters.

This feature is useful for adaptive SCHED_DEADLINE tasks that need to
modify their behavior depending on whether or not there is enough runtime
left in the current period, and/or what is the current absolute deadline.

Notes:
- before returning the instantaneous parameters, the runtime is updated;
- the abs deadline is returned shifted from rq_clock() to ktime_get_ns(),
  in CLOCK_MONOTONIC reference; this causes multiple invocations from the
  same period to return values that may differ for a few ns (showing some
  small drift), albeit the deadline doesn't move, in rq_clock() reference;
- the abs deadline value returned to user-space, as unsigned 64-bit value,
  can represent nearly 585 years since boot time;
- setting flags=0 provides the old behavior (retrieve static parameters).

See also the notes from discussion held at OSPM 2025 on the topic
"Making user space aware of current deadline-scheduler parameters".

Signed-off-by: Tommaso Cucinotta <tommaso.cucinotta@santannapisa.it>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Matteo Martelli <matteo.martelli@codethink.co.uk>
Link: https://patch.msgid.link/20250912053937.31636-2-tommaso.cucinotta@santannapisa.it
2026-02-25 15:35:58 +01:00
Peter Zijlstra
c9bc1753b3 perf: Fix __perf_event_overflow() vs perf_remove_from_context() race
Make sure that __perf_event_overflow() runs with IRQs disabled for all
possible callchains. Specifically the software events can end up running
it with only preemption disabled.

This opens up a race vs perf_event_exit_event() and friends that will go
and free various things the overflow path expects to be present, like
the BPF program.

Fixes: 592903cdcb ("perf_counter: add an event_list")
Reported-by: Simond Hu <cmdhh1767@gmail.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Simond Hu <cmdhh1767@gmail.com>
Link: https://patch.msgid.link/20260224122909.GV1395416@noisy.programming.kicks-ass.net
2026-02-25 15:02:34 +01:00
Mark Brown
463279e588
spi: add devm_spi_new_ancillary_device()
Merge series from Antoniu Miclaus <antoniu.miclaus@analog.com>:

Add a devres-managed version of spi_new_ancillary_device() that
automatically unregisters the ancillary SPI device when the parent
device is removed.
2026-02-25 13:42:27 +00:00
Srinivas Pandruvada
6b050482ec cpufreq: intel_pstate: Fix crash during turbo disable
When the system is booted with kernel command line argument "nosmt" or
"maxcpus" to limit the number of CPUs, disabling turbo via:

 echo 1 > /sys/devices/system/cpu/intel_pstate/no_turbo

results in a crash:

 PF: supervisor read access in kernel mode
 PF: error_code(0x0000) - not-present page
 PGD 0 P4D 0
 Oops: Oops: 0000 [#1] SMP PTI
 ...
 RIP: 0010:store_no_turbo+0x100/0x1f0
 ...

This occurs because for_each_possible_cpu() returns CPUs even if they
are not online. For those CPUs, all_cpu_data[] will be NULL. Since
commit 973207ae3d ("cpufreq: intel_pstate: Rearrange max frequency
updates handling code"), all_cpu_data[] is dereferenced even for CPUs
which are not online, causing the NULL pointer dereference.

To fix that, pass CPU number to intel_pstate_update_max_freq() and use
all_cpu_data[] for those CPUs for which there is a valid cpufreq policy.

Fixes: 973207ae3d ("cpufreq: intel_pstate: Rearrange max frequency updates handling code")
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=221068
Signed-off-by: Srinivas Pandruvada <srinivas.pandruvada@linux.intel.com>
Cc: 6.16+ <stable@vger.kernel.org> # 6.16+
Link: https://patch.msgid.link/20260225001752.890164-1-srinivas.pandruvada@linux.intel.com
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
2026-02-25 14:39:19 +01:00
Imre Deak
230f8c9956 drm/i915/dp_tunnel: Send BW change notification after tunnel creation
Detecting a bandwidth change for a sink connected through a DP tunnel
depends on updating the sink's DPRX link rate and lane count.

detect_new_tunnel() -> update_tunnel_state() updates the link
configuration only if the tunnel state changes. However, after the
tunnel is created and bandwidth allocation mode is enabled, the tunnel
state itself may remain unchanged.

Record the sink bandwidth before creating the tunnel and compare it to
the bandwidth after tunnel creation and enabling bandwidth allocation
mode, ensuring that any bandwidth change is detected and userspace is
notified accordingly.

Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patch.msgid.link/20260219182823.926702-6-imre.deak@intel.com
2026-02-25 15:30:23 +02:00
Imre Deak
23a5fd5bec drm/i915/dp_tunnel: Sanitize documentation of intel_dp_tunnel_detect()
Clarify the documentation of detect_new_tunnel() return values,
including the failure case.

Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patch.msgid.link/20260219182823.926702-5-imre.deak@intel.com
2026-02-25 15:30:19 +02:00
Imre Deak
bcc7907398 drm/i915/dp_tunnel: Split update_tunnel_state()
Split update_tunnel_state() into two helpers: one that updates the
tunnel state, and another that detects whether the tunnel bandwidth
has changed.

This prepares for a follow-up change that needs to compare the current
bandwidth against the value from before the DP tunnel was detected and
bandwidth allocation mode was enabled.

While at it, document the return value of update_tunnel_state().

Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patch.msgid.link/20260219182823.926702-4-imre.deak@intel.com
2026-02-25 15:30:16 +02:00
Imre Deak
02feab8237 drm/i915/dp_tunnel: Simplify detection of link BW change
update_tunnel_state() checks whether a tunnel state change (e.g.
available tunnel bandwidth) affects the list of valid modes for the
sink connected through the tunnel. If so, its caller sends a hotplug
event so userspace can re-enumerate the modes.

A change in tunnel bandwidth does not affect the mode list if the
bandwidth was above the sink's DPRX bandwidth both before and after the
update, since in that case the effective bandwidth remains limited by
the DPRX.

As get_current_link_bw() via intel_dp_max_link_data_rate() already
returns bandwidth values clamped to the DPRX limit, the condition for
detecting a mode list change can be simplified to:

    old_bw != new_bw

Remove the explicit checks for whether the bandwidth was below the
maximum DPRX bandwidth before and after the update, and rely on the
clamped bandwidth values instead.

Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patch.msgid.link/20260219182823.926702-3-imre.deak@intel.com
2026-02-25 15:30:11 +02:00
Imre Deak
afe3f74716 drm/i915/dp_tunnel: Don't update tunnel state during system resume
During system resume, restoring the pre-suspend display state must not
fail. This requires preserving the sink capabilities from before
suspend, including the available link bandwidth.

If these capabilities are not preserved, the restore modeset may fail,
either due to a missing sink capability or insufficient link bandwidth
for the restored mode.

When the sink is connected through a DP tunnel, prevent such capability
changes by skipping tunnel state updates during resume. This also avoids
updating the sink state via the tunnel while it is being resumed.

Reviewed-by: Arun R Murthy <arun.r.murthy@intel.com>
Signed-off-by: Imre Deak <imre.deak@intel.com>
Link: https://patch.msgid.link/20260219182823.926702-2-imre.deak@intel.com
2026-02-25 15:29:46 +02:00
Leon Romanovsky
58409f0d4d RDMA/mlx4: Remove unused create_flags field from CQ structure
The CQ creation flags do not need to be cached, as they are used
immediately at the point where they are stored. Remove the unused
field and reclaim 4 bytes.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-14-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
f45f195af5 RDMA/mlx4: Introduce a modern CQ creation interface
The uverbs CQ creation UAPI allows users to supply their own umem when
creating a CQ. Update mlx4 to support this model while preserving compatibility
with the legacy interface that allocates umem internally.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-13-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
0e4b9841f4 RDMA/mlx4: Inline mlx4_ib_get_cq_umem into callers
Inline the mlx4_ib_get_cq_umem helper function into its two call sites
(mlx4_ib_create_cq and mlx4_alloc_resize_umem) to prepare for the
transition to modern CQ creation interface.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-12-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
14738dce7e RDMA/mlx5: Provide a modern CQ creation interface
The uverbs CQ creation UAPI allows users to supply their own umem for a CQ.
Update mlx5 to support this workflow while preserving support for creating
umem through the legacy interface.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-11-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
eebea464f5 RDMA/mlx5: Save 4 bytes in CQ structure
There is no need to maintain separate, nearly empty create_flags and
private_flags fields. Unifying them reduces memory usage.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-10-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
66011c1bd7 RDMA/efa: Remove check for zero CQE count
Since ib_core now handles validation, the device driver no longer needs
to verify that the CQE count is non‑zero.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-9-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
a291758288 RDMA/core: Reject zero CQE count
All drivers already ensure that the number of CQEs is at least 1.
Add this validation to the core so drivers no longer need to repeat it.
Future patches converting to the .create_user_cq() interface will remove
the per‑driver checks.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-8-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
584ec74748 RDMA/core: Prepare create CQ path for API unification
Ensure that .create_cq_umem() and .create_cq() follow the same API
contract, allowing drivers to be gradually migrated to the umem-aware
CQ management flow.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-7-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
2ead7b09bc RDMA/efa: Rely on CPU address in create‑QP
Align this code with other locations where efa_free_mapped() depends on the
presence of a valid CPU address, which is guaranteed when qp->rq_size != 0.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-6-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
25c7410488 RDMA/core: Manage CQ umem in core code
In the current implementation, CQ umem is handled both by ib_core and
the driver. ib_core sometimes creates and destroys it, while the driver
also destroys it.

Store the umem in struct ib_cq and ensure that only ib_core manages
its lifetime, relying solely on its internal reference counter.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-5-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
a731c86265 RDMA/core: Promote UMEM to a core component
To manage UMEM objects at the core level and reuse the existing
ib_destroy_cq*() flow, move the UMEM files to be built together with
ib_core. Attempting to call ib_umem_release() from verbs.c currently
results in the following error:

    depmod: ERROR: Cycle detected: ib_core -> ib_uverbs -> ib_core
    depmod: ERROR: Found 2 modules in dependency cycles!
    verbs.c:(.text+0x250c): undefined reference to `ib_umem_release'

Link: https://patch.msgid.link/20260213-refactor-umem-v1-4-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
e3104fe921 RDMA/umem: Remove unnecessary includes and defines from ib_umem header
The ib_umem header no longer requires the removed includes or forward
declarations, so drop them to reduce clutter.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-3-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
2ae3c4f6ea RDMA/umem: Allow including ib_umem header from any location
Including ib_umem.h currently triggers circular dependency errors.
These issues can be resolved by removing the include of ib_verbs.h,
which was only needed to resolve the struct ib_device pointer.

>> depmod: ERROR: Cycle detected: ib_core -> ib_uverbs -> ib_core
>> depmod: ERROR: Found 2 modules in dependency cycles!
  make[3]: *** [scripts/Makefile.modinst:132: depmod] Error 1
  make[3]: Target '__modinst' not remade because of errors.
  make[2]: *** [Makefile:1960: modules_install] Error 2
  make[1]: *** [Makefile:248: __sub-make] Error 2
  make[1]: Target 'modules_install' not remade because of errors.
  make: *** [Makefile:248: __sub-make] Error 2
  make: Target 'modules_install' not remade because of errors.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-2-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Leon Romanovsky
6094ea64c6 RDMA: Move DMA block iterator logic into dedicated files
The DMA iterator logic was mixed into verbs and umem-specific code,
forcing all users to include rdma/ib_umem.h. Move the block iterator
logic into iter.c and rdma/iter.h so that rdma/ib_umem.h and
rdma/ib_verbs.h can be separated in a follow-up patch.

Link: https://patch.msgid.link/20260213-refactor-umem-v1-1-f3be85847922@nvidia.com
Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
2026-02-25 08:15:30 -05:00
Wilfred Mallawa
650b774cf9 xfs: add static size checks for ioctl UABI
The ioctl structures in libxfs/xfs_fs.h are missing static size checks.
It is useful to have static size checks for these structures as adding
new fields to them could cause issues (e.g. extra padding that may be
inserted by the compiler). So add these checks to xfs/xfs_ondisk.h.

Due to different padding/alignment requirements across different
architectures, to avoid build failures, some structures are ommited from
the size checks. For example, structures with "compat_" definitions in
xfs/xfs_ioctl32.h are ommited.

Signed-off-by: Wilfred Mallawa <wilfred.mallawa@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:50 +01:00
Wilfred Mallawa
e97cbf863d xfs: remove duplicate static size checks
In libxfs/xfs_ondisk.h, remove some duplicate entries of
XFS_CHECK_STRUCT_SIZE().

Signed-off-by: Wilfred Mallawa <wilfred.mallawa@wdc.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00
Nirjhar Roy (IBM)
9a654a8fa3 xfs: Add comments for usages of some macros.
Add comments explaining when to use XFS_IS_CORRUPT() and ASSERT()

Suggested-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nirjhar Roy (IBM) <nirjhar.roy.lists@gmail.com>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00
Nirjhar Roy (IBM)
c2368fc89a xfs: Update lazy counters in xfs_growfs_rt_bmblock()
Update lazy counters in xfs_growfs_rt_bmblock() similar to the way it
is done xfs_growfs_data_private(). This is because the lazy counters are
not always updated and synching the counters will avoid inconsistencies
between frexents and rtextents(total realtime extent count). This will
be more useful once realtime shrink is implemented as this will prevent
some transient state to occur where frexents might be greater than
total rtextents.

Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nirjhar Roy (IBM) <nirjhar.roy.lists@gmail.com>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00
Nirjhar Roy (IBM)
ac1d977096 xfs: Add a comment in xfs_log_sb()
Add a comment explaining why the sb_frextents are updated outside the
if (xfs_has_lazycount(mp) check even though it is a lazycounter.
RT groups are supported only in v5 filesystems which always have
lazycounter enabled - so putting it inside the if(xfs_has_lazycount(mp)
check is redundant.

Suggested-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Nirjhar Roy (IBM) <nirjhar.roy.lists@gmail.com>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00
Nirjhar Roy (IBM)
8baa9bccc0 xfs: Fix xfs_last_rt_bmblock()
Bug description:

If the size of the last rtgroup i.e, the rtg passed to
xfs_last_rt_bmblock() is such that the last rtextent falls in 0th word
offset of a bmblock of the bitmap file tracking this (last) rtgroup,
then in that case xfs_last_rt_bmblock() incorrectly returns the next
bmblock number instead of the current/last used bmblock number.
When xfs_last_rt_bmblock() incorrectly returns the next bmblock,
the loop to grow/modify the bmblocks in xfs_growfs_rtg() doesn't
execute and xfs_growfs basically does a nop in certain cases.

xfs_growfs will do a nop when the new size of the fs will have the same
number of rtgroups i.e, we are only growing the last rtgroup.

Reproduce:
$ mkfs.xfs -m metadir=0 -r rtdev=/dev/loop1 /dev/loop0 \
	-r size=32769b -f
$ mount -o rtdev=/dev/loop1 /dev/loop0 /mnt/scratch
$ xfs_growfs -R $(( 32769 + 1 )) /mnt/scratch
$ xfs_info /mnt/scratch | grep rtextents
$ # We can see that rtextents hasn't changed

Fix:
Fix this by returning the current/last used bmblock when the last
rtgroup size is not a multiple xfs_rtbitmap_rtx_per_rbmblock()
and the next bmblock when the rtgroup size is a multiple of
xfs_rtbitmap_rtx_per_rbmblock() i.e, the existing blocks are
completely used up.
Also, I have renamed xfs_last_rt_bmblock() to
xfs_last_rt_bmblock_to_extend() to signify that this function
returns the bmblock number to extend and NOT always the last used
bmblock number.

Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Nirjhar Roy (IBM) <nirjhar.roy.lists@gmail.com>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00
Darrick J. Wong
115ea07b94 xfs: don't report half-built inodes to fserror
Sam Sun apparently found a syzbot way to fuzz a filesystem such that
xfs_iget_cache_miss would free the inode before the fserror code could
catch up.  Frustratingly he doesn't use the syzbot dashboard so there's
no C reproducer and not even a full error report, so I'm guessing that:

Inodes that are being constructed or torn down inside XFS are not
visible to the VFS.  They should never be reported to fserror.
Also, any inode that has been freshly allocated in _cache_miss should be
marked INEW immediately because, well, it's an incompletely constructed
inode that isn't yet visible to the VFS.

Reported-by: Sam Sun <samsun1006219@gmail.com>
Fixes: 5eb4cb18e4 ("xfs: convey metadata health events to the health monitor")
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00
Darrick J. Wong
75690e5fdd xfs: don't report metadata inodes to fserror
Internal metadata inodes are not exposed to userspace programs, so it
makes no sense to pass them to the fserror functions (aka fsnotify).
Instead, report metadata file problems as general filesystem corruption.

Fixes: 5eb4cb18e4 ("xfs: convey metadata health events to the health monitor")
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00
Darrick J. Wong
94014a23e9 xfs: fix potential pointer access race in xfs_healthmon_get
Pankaj Raghav asks about this code in xfs_healthmon_get:

  hm = mp->m_healthmon;
  if (hm && !refcount_inc_not_zero(&hm->ref))
    hm = NULL;
  rcu_read_unlock();
  return hm;

(slightly edited to compress a mailing list thread)

"Nit: Should we do a READ_ONCE(mp->m_healthmon) here to avoid any
compiler tricks that can result in an undefined behaviour? I am not sure
if I am being paranoid here.

"So this is my understanding: RCU guarantees that we get a valid object
(actual data of m_healthmon) but does not guarantee the compiler will
not reread the pointer between checking if hm is !NULL and accessing the
pointer as we are doing it lockless.

"So just a barrier() call in rcu_read_lock is enough to make sure this
doesn't happen and probably adding a READ_ONCE() is not needed?"

After some initial confusion I concluded that he's correct.  The
compiler could very well eliminate the hm variable in favor of walking
the pointers twice, turning the code into:

  if (mp->m_healthmon && !refcount_inc_not_zero(&mp->m_healthmon->ref))

If this happens, then xfs_healthmon_detach can sneak in between the
two sides of the && expression and set mp->m_healthmon to NULL, and
thereby cause a null pointer dereference crash.  Fix this by using the
rcu pointer assignment and dereference functions, which ensure that the
proper reordering barriers are in place.

Practically speaking, gcc seems to allocate an actual variable for hm
and only reads mp->m_healthmon once (as intended), but we ought to be
more explicit about requiring this.

Reported-by: Pankaj Raghav <pankaj.raghav@linux.dev>
Fixes: a48373e7d3 ("xfs: start creating infrastructure for health monitoring")
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Reviewed-by: Pankaj Raghav <p.raghav@samsung.com>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00
Darrick J. Wong
eb8550fb75 xfs: fix xfs_group release bug in xfs_dax_notify_dev_failure
Chris Mason reports that his AI tools noticed that we were using
xfs_perag_put and xfs_group_put to release the group reference returned
by xfs_group_next_range.  However, the iterator function returns an
object with an active refcount, which means that we must use the correct
function to release the active refcount, which is _rele.

Cc: <stable@vger.kernel.org> # v6.0
Fixes: 6f643c57d5 ("xfs: implement ->notify_failure() for XFS")
Signed-off-by: "Darrick J. Wong" <djwong@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com>
Signed-off-by: Carlos Maiolino <cem@kernel.org>
2026-02-25 13:58:49 +01:00