phy: qcom: phy-qcom-m31: Update IPQ5332 M31 USB phy initialization sequence
The current configuration used for the IPQ5332 M31 USB PHY fails the
Near End High Speed Signal Quality compliance test. To resolve this,
update the initialization sequence as specified in the Hardware Design
Document.
dt-bindings: phy: Convert ti,da830-usb-phy to DT schema
Convert the TI DA830 USB PHY binding to DT schema format. Add "clocks"
and "clock-names" which are already in use. As they are always present,
make them required as well.
Signed-off-by: Rob Herring (Arm) <robh@kernel.org> Reviewed-by: David Lechner <david@lechnology.com> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/20250627220107.214162-1-robh@kernel.org Signed-off-by: Vinod Koul <vkoul@kernel.org>
phy: exynos-mipi-video: correct cam0 sysreg property name for exynos7870
Fix the cam0 sysreg property name (samsung,cam0-sysreg), which has been
erroneously declared as samsung,cam-sysreg. This follows the same name
used in Exynos5433 PHY.
Mingming Cao [Sat, 19 Jul 2025 09:13:56 +0000 (05:13 -0400)]
ibmveth: Add multi buffers rx replenishment hcall support
This patch enables batched RX buffer replenishment in ibmveth by
using the new firmware-supported h_add_logical_lan_buffers() hcall
to submit up to 8 RX buffers in a single call, instead of repeatedly
calling the single-buffer h_add_logical_lan_buffer() hcall.
During the probe, with the patch, the driver queries ILLAN attributes
to detect IBMVETH_ILLAN_RX_MULTI_BUFF_SUPPORT bit. If the attribute is
present, rx_buffers_per_hcall is set to 8, enabling batched replenishment.
Otherwise, it defaults to 1, preserving the original upstream behavior
with no change in code flow for unsupported systems.
The core rx replenish logic remains the same. But when batching
is enabled, the driver aggregates up to 8 fully prepared descriptors
into a single h_add_logical_lan_buffers() hypercall. If any allocation
or DMA mapping fails while preparing a batch, only the successfully
prepared buffers are submitted, and the remaining are deferred for
the next replenish cycle.
If at runtime the firmware stops accepting the batched hcall—e,g,
after a Live Partition Migration (LPM) to a host that does not
support h_add_logical_lan_buffers(), the hypercall returns H_FUNCTION.
In that case, the driver transparently disables batching, resets
rx_buffers_per_hcall to 1, and falls back to the single-buffer hcall
in next future replenishments to take care of these and future buffers.
Test were done on systems with firmware that both supports and
does not support the new h_add_logical_lan_buffers hcall.
On supported firmware, this reduces hypercall overhead significantly
over multiple buffers. SAR measurements showed about a 15% improvement
in packet processing rate under moderate RX load, with heavier traffic
seeing gains more than 30%
Signed-off-by: Mingming Cao <mmc@linux.ibm.com> Reviewed-by: Brian King <bjking1@linux.ibm.com> Reviewed-by: Haren Myneni <haren@linux.ibm.com> Reviewed-by: Dave Marquardt <davemarq@linux.ibm.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20250719091356.57252-1-mmc@linux.ibm.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
platform/x86/amd: pmc: Add Lenovo Yoga 6 13ALC6 to pmc quirk list
The Lenovo Yoga 6 13ACL6 82ND has a similar BIOS problem as other Lenovo
laptops from that vintage that causes a rather long resume from suspend.
Add it to the quirk list that manipulates the scratch register to avoid
the issue.
Reported-by: Adam Berglund <adam.f.berglund@hotmail.com> Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/4434 Tested-by: Adam Berglund <adam.f.berglund@hotmail.com> Signed-off-by: Mario Limonciello <mario.limonciello@amd.com> Reviewed-by: Hans de Goede <hansg@kernel.org> Link: https://lore.kernel.org/r/20250718172307.1928744-1-superm1@kernel.org Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
platform/x86: dell-uart-backlight: Use blacklight power constant
The backlight subsystem has gotten its own power constants. Replace
FB_BLANK_UNBLANK with BACKLIGHT_POWER_ON.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de> Reviewed-by: Hans de Goede <hansg@kernel.org> Link: https://lore.kernel.org/r/20250715122643.137027-2-tzimmermann@suse.de Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com> Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Merge tag 'thermal-v6.17-rc1' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/thermal/linux
Merge thermal driver updates for 6.17-rc1 from Daniel Lezcano:
"- Enable the stage 2 shutdown and support to more SPMI variants
(Anjelique Melendez)
- Constify thermal_zone_device_ops structure when possible in the
different thermal drivers (Christophe Jaillet)
- Use the dev_fwnode() helper instead of of_fwnode_handle() which is
more adequate wherever is possible in the thermal drivers (Jiri
Slaby)
- Implement and document One-Time Programmable fuse support for the
Rockchip driver in order to increase the precision of the
measurements (Nicolas Frattaroli)
- Change the way the Mediatek LTVS driver stores the initialization
data sequence to support different sequences regarding the current
platform. Introduce the mt7988 support with a new initialization
sequence (Mason Chang)
- Document the QCom TSens Milos Temperature Sensor DT bindings (Luca
Weiss)
- Add the fallback compatible string for MT7981 and MT8516 DT bindings
(Aleksander Jan Bajkowski)
- Add the compatible string for Tegra210B01 SOC_THERM driver (Aaron Kling)"
* tag 'thermal-v6.17-rc1' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/thermal/linux:
dt-bindings: thermal: tegra: Document Tegra210B01
dt-bindings: thermal: mediatek: Add fallback compatible string for MT7981 and MT8516
dt-bindings: thermal: qcom-tsens: document the Milos Temperature Sensor
thermal/drivers/mediatek/lvts_thermal: Add mt7988 lvts commands
thermal/drivers/mediatek/lvts_thermal: Add lvts commands and their sizes to driver data
thermal/drivers/mediatek/lvts_thermal: Change lvts commands array to static const
thermal/drivers/rockchip: Support reading trim values from OTP
dt-bindings: thermal: rockchip: document otp thermal trim
thermal/drivers/rockchip: Support RK3576 SoC in the thermal driver
dt-bindings: rockchip-thermal: Add RK3576 compatible
thermal/drivers/rockchip: Rename rk_tsadcv3_tshut_mode
thermal: Use dev_fwnode()
thermal: Constify struct thermal_zone_device_ops
thermal/drivers/loongson2: Constify struct thermal_zone_device_ops
thermal/drivers/qcom-spmi-temp-alarm: Add support for LITE PMIC peripherals
thermal/drivers/qcom-spmi-temp-alarm: Add support for GEN2 rev 2 PMIC peripherals
thermal/drivers/qcom-spmi-temp-alarm: Prepare to support additional Temp Alarm subtypes
thermal/drivers/qcom-spmi-temp-alarm: Add temp alarm data struct based on HW subtype
thermal/drivers/qcom-spmi-temp-alarm: Enable stage 2 shutdown when required
phy: qcom: phy-qcom-snps-eusb2: Add missing write from init sequence
As per a commit from Qualcomm's downstream 6.1 kernel[0], the init
sequence is missing setting the CMN_CTRL_OVERRIDE_EN bit back to 0 at
the end, as per the 'latest' HPG revision (as of November 2023).
Thomas Gleixner [Fri, 18 Jul 2025 18:54:12 +0000 (20:54 +0200)]
genirq: Prevent migration live lock in handle_edge_irq()
Yicon reported and Liangyan debugged a live lock in handle_edge_irq()
related to interrupt migration.
If the interrupt affinity is moved to a new target CPU and the interrupt is
currently handled on the previous target CPU for edge type interrupts the
handler might get stuck on the previous target:
CPU 0 (previous target) CPU 1 (new target)
handle_edge_irq()
repeat:
handle_event() handle_edge_irq()
if (INPROGESS) {
set(PENDING);
mask();
return;
}
if (PENDING) {
clear(PENDING);
unmask();
goto repeat;
}
The migration in software never completes and CPU0 continues to handle the
pending events forever. This happens when the device raises interrupts with
a high rate and always before handle_event() completes and before the CPU0
handler can clear INPROGRESS so that CPU1 sets the PENDING flag over and
over. This has been observed in virtual machines.
Prevent this by checking whether the CPU which observes the INPROGRESS flag
is the new affinity target. If that's the case, do not set the PENDING flag
and wait for the INPROGRESS flag to be cleared instead, so that the new
interrupt is handled on the new target CPU and the previous CPU is released
from the action.
This is restricted to the edge type handler and only utilized on systems,
which use single CPU targets for interrupt affinity.
Thomas Gleixner [Fri, 18 Jul 2025 18:54:08 +0000 (20:54 +0200)]
genirq: Move irq_wait_for_poll() to call site
Move it to the call site so that the waiting for the INPROGRESS flag can be
reused by an upcoming mitigation for a potential live lock in the edge type
handler.
phy: rockchip-pcie: Properly disable TEST_WRITE strobe signal
pcie_conf is used to touch TEST_WRITE strobe signal. This signal should
be enabled, a little time waited, and then disabled. Current code clearly
was copy-pasted and never disables the strobe signal. Adjust the define.
While at it, remove PHY_CFG_RD_MASK which has been unused since 64cdc0360811 ("phy: rockchip-pcie: remove unused phy_rd_cfg function").
phy: rockchip-pcie: Enable all four lanes if required
Current code enables only Lane 0 because pwr_cnt will be incremented on
first call to the function. Let's reorder the enablement code to enable
all 4 lanes through GRF.
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org> Reviewed-by: Robin Murphy <robin.murphy@arm.com> Signed-off-by: Valmantas Paliksa <walmis@gmail.com> Signed-off-by: Geraldo Nascimento <geraldogabriel@gmail.com> Reviewed-by: Robin Murphy <robin.murphy@arm.com> Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org> Link: https://lore.kernel.org/r/16b610aab34e069fd31d9f57260c10df2a968f80.1751322015.git.geraldogabriel@gmail.com Signed-off-by: Vinod Koul <vkoul@kernel.org>
Markus Blöchl [Sun, 20 Jul 2025 13:54:51 +0000 (15:54 +0200)]
timekeeping: Zero initialize system_counterval when querying time from phc drivers
Most drivers only populate the fields cycles and cs_id of system_counterval
in their get_time_fn() callback for get_device_system_crosststamp(), unless
they explicitly provide nanosecond values.
When the use_nsecs field was added to struct system_counterval, most
drivers did not care. Clock sources other than CSID_GENERIC could then get
converted in convert_base_to_cs() based on an uninitialized use_nsecs field,
which usually results in -EINVAL during the following range check.
Pass in a fully zero initialized system_counterval_t to cure that.
dt-bindings: phy: qcom,sc8280xp-qmp-pcie-phy: Update pcie phy bindings for QCS615
QCS615 pcie phy only use 5 clocks, which are aux, cfg_ahb, ref,
ref_gen, pipe. So move "qcom,qcs615-qmp-gen3x1-pcie-phy" compatible
from 6 clocks' list to 5 clocks' list.
In ieee80211_latest_active_link_conn_timeout() we loop over all
sta->links in order to compute the timeout expiring last across
all links.
Such timeout is stored in `latest_timeout` which is used in the
time_after() comparison before having been initialized.
Fix this behaviour by initializing the variable to `jiffies` and
adapt surrouding conditions accordingly.
Note that the caller assumed latest_timeout to be 0 if no active
link was found. This is not appropriate because jiffies=0 is a
valid (and recurrent, although not often) point in time.
By using `jiffies` as default value for latest_timeout, we can
fix the caller as well.
Address-Coverity-ID: 1647986 ("Uninitialized variables (UNINIT)") Fixes: 1bc892d76a6f ("wifi: mac80211: extend connection monitoring for MLO") Signed-off-by: Antonio Quartulli <antonio@mandelbit.com> Link: https://patch.msgid.link/20250722120634.3501-1-antonio@mandelbit.com Signed-off-by: Johannes Berg <johannes.berg@intel.com>
SUNRPC: Silence warnings about parameters not being described
Warning: net/sunrpc/auth_gss/gss_krb5_crypto.c:902 function parameter
'len' not described in 'krb5_etm_decrypt'
Warning: net/sunrpc/auth_gss/gss_krb5_crypto.c:902 function parameter
'buf' not described in 'krb5_etm_decrypt'
nfs_delegation_find_inode currently has to walk the entire list of
delegations per inode, which can become pretty large, and can become even
larger when increasing the delegation watermark.
Add a hash table to speed up the delegation lookup, sized as a fraction
of the delegation watermark.
The active delegation watermark was added to avoid overloading servers.
Track the active delegation per-server instead of globally so that clients
talking to multiple servers aren't limited by the global limit.
Tigran Mkrtchyan [Fri, 27 Jun 2025 07:17:51 +0000 (09:17 +0200)]
pNFS/flexfiles: don't attempt pnfs on fatal DS errors
When an applications get killed (SIGTERM/SIGINT) while pNFS client performs a connection
to DS, client ends in an infinite loop of connect-disconnect. This
source of the issue, it that flexfilelayoutdev#nfs4_ff_layout_prepare_ds gets an error
on nfs4_pnfs_ds_connect with status ERESTARTSYS, which is set by rpc_signal_task, but
the error is treated as transient, thus retried.
The issue is reproducible with Ctrl+C the following script(there should be ~1000 files in
a directory, client should must not have any connections to DSes):
```
echo 3 > /proc/sys/vm/drop_caches
for i in *
do
head -1 $i
done
```
The change aims to propagate the nfs4_ff_layout_prepare_ds error state
to the caller that can decide whatever this is a retryable error or not.
My previous patch ended up causing a regression for the
DRM_IOCTL_NOUVEAU_NVIF ioctl. The intention of my patch was to only
pass ioctl commands that have the correct dir/type/nr bits into the
nouveau_abi16_ioctl() function.
This turned out to be too strict, as userspace does use at least
write-only and write-read direction settings. Checking for both of these
still did not fix the issue, so the best we can do for the 6.16 release
is to revert back to what we've had since linux-3.16.
This version is still fragile, but at least it is known to work with
existing userspace. Fixing this properly requires a better understanding
of what commands are being passed from userspace in practice, and how
that relies on the undocumented (miss)behavior in nouveau_drm_ioctl().
rust: kernel: move ARef and AlwaysRefCounted to sync::aref
Move the definitions of `ARef` and `AlwaysRefCounted` from `types.rs`
to a new file `sync/aref.rs`. Define the corresponding `aref` module
under `rust/kernel/sync.rs`. These types are better grouped in `sync`.
To avoid breaking existing imports, they are re-exported from `types.rs`.
Drop unused imports `mem::ManuallyDrop`, `ptr::NonNull` from `types.rs`,
they are now only used in `sync/aref.rs`, where they are already imported.
Benno Lossin [Tue, 20 May 2025 23:17:13 +0000 (01:17 +0200)]
rust: sync: fix safety comment for `static_lock_class`
The safety comment mentions lockdep -- which from a Rust perspective
isn't important -- and doesn't mention the real reason for why it's
sound to create `LockClassKey` as uninitialized memory.
Joel Fernandes [Tue, 15 Jul 2025 20:01:55 +0000 (16:01 -0400)]
rcu: Document concurrent quiescent state reporting for offline CPUs
The synchronization of CPU offlining with GP initialization is confusing
to put it mildly (rightfully so as the issue it deals with is complex).
Recent discussions brought up a question -- what prevents the
rcu_implicit_dyntick_qs() from warning about QS reports for offline
CPUs (missing QS reports for offline CPUs causing indefinite hangs).
QS reporting for now-offline CPUs should only happen from:
- gp_init()
- rcutree_cpu_report_dead()
Add some documentation on this and refer to it from comments in the code
explaining how QS reporting is not missed when these functions are
concurrently running.
I referred heavily to this post [1] about the need for the ofl_lock.
[1] https://lore.kernel.org/all/20180924164443.GF4222@linux.ibm.com/
[ Applied paulmck feedback on moving documentation to Requirements.rst ]
Link: https://lore.kernel.org/all/01b4d228-9416-43f8-a62e-124b92e8741a@paulmck-laptop/ Co-developed-by: "Paul E. McKenney" <paulmck@kernel.org> Signed-off-by: "Paul E. McKenney" <paulmck@kernel.org> Reviewed-by: Frederic Weisbecker <frederic@kernel.org> Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com> Signed-off-by: Neeraj Upadhyay (AMD) <neeraj.upadhyay@kernel.org>
Add driver for the Qualcomm interconnect buses found in Milos based
platforms. The topology consists of several NoCs that are controlled by
a remote processor that collects the aggregated bandwidth for each
master-slave pairs.
Stephen Rothwell [Tue, 22 Jul 2025 10:27:54 +0000 (11:27 +0100)]
ASoC: SDCA: Check devm_mutex_init() return value
Fix interaction with commit daec29dcc873 ("locking/mutex: Mark
devm_mutex_init() as __must_check"), add return value check. There is no
need for additional complex error handling here, failure to init the
mutex means the code can't progress, so the failure just needs to be passed
up to the caller.
Andy Shevchenko [Wed, 2 Jul 2025 06:15:34 +0000 (09:15 +0300)]
rtc: sysfs: Use sysfs_emit() to instead of s*printf()
Follow the advice of the Documentation/filesystems/sysfs.rst that show()
should only use sysfs_emit() or sysfs_emit_at() when formatting the value
to be returned to user space.
Convert the Renesas SuperH On-Chip RTC driver from SIMPLE_DEV_PM_OPS()
to DEFINE_SIMPLE_DEV_PM_OPS() and pm_sleep_ptr(). This lets us drop the
__maybe_unused annotations from its suspend and resume callbacks, and
reduces kernel size in case CONFIG_PM or CONFIG_PM_SLEEP is disabled.
When compiling the RTC library functions test as a module, the module
has the non-descriptive name "lib_test.ko". Fix this by renaming it to
"test_rtc_lib.ko".
can: netlink: can_changelink(): fix NULL pointer deref of struct can_priv::do_set_mode
Andrei Lalaev reported a NULL pointer deref when a CAN device is
restarted from Bus Off and the driver does not implement the struct
can_priv::do_set_mode callback.
There are 2 code path that call struct can_priv::do_set_mode:
- directly by a manual restart from the user space, via
can_changelink()
- delayed automatic restart after bus off (deactivated by default)
To prevent the NULL pointer deference, refuse a manual restart or
configure the automatic restart delay in can_changelink() and report
the error via extack to user space.
As an additional safety measure let can_restart() return an error if
can_priv::do_set_mode is not set instead of dereferencing it
unchecked.
Merge tag 'md-6.17-20250722' of https://git.kernel.org/pub/scm/linux/kernel/git/mdraid/linux into for-6.17/block
Pull MD updates from Yu:
"- call del_gendisk synchronously, from Xiao
- cleanup unused variable, from John
- cleanup workqueue flags, from Ryo
- fix faulty rdev can't be removed during resync, from Qixing"
* tag 'md-6.17-20250722' of https://git.kernel.org/pub/scm/linux/kernel/git/mdraid/linux:
md/raid10: fix set but not used variable in sync_request_write()
md: allow removing faulty rdev during resync
md/raid5: unset WQ_CPU_INTENSIVE for raid5 unbound workqueue
md: remove/add redundancy group only in level change
md: Don't clear MD_CLOSING until mddev is freed
md: call del_gendisk in control path
Merge tag 'nvme-6.17-2025-07-22' of git://git.infradead.org/nvme into for-6.17/block
Pull NVMe updates from Christoph:
"- try PCIe function level reset on init failure (Keith Busch)
- log TLS handshake failures at error level (Maurizio Lombardi)
- pci-epf: do not complete commands twice if nvmet_req_init() fails
(Rick Wertenbroek)
- misc cleanups (Alok Tiwari)"
* tag 'nvme-6.17-2025-07-22' of git://git.infradead.org/nvme:
nvme-pci: try function level reset on init failure
nvmet: pci-epf: Do not complete commands twice if nvmet_req_init() fails
nvme-tcp: log TLS handshake failures at error level
docs: nvme: fix grammar in nvme-pci-endpoint-target.rst
nvme: fix typo in status code constant for self-test in progress
nvmet: remove redundant assignment of error code in nvmet_ns_enable()
nvme: fix incorrect variable in io cqes error message
nvme: fix multiple spelling and grammar issues in host drivers
cpufreq: CPPC: Mark driver with NEED_UPDATE_LIMITS flag
AMU counters on certain CPPC-based platforms tend to yield inaccurate
delivered performance measurements on systems that are idle/mostly idle.
This results in an inaccurate frequency being stored by cpufreq in its
policy structure when the CPU is brought online. [1]
Consequently, if the userspace governor tries to set the frequency to a
new value, there is a possibility that it would be the erroneous value
stored earlier. In such a scenario, cpufreq would assume that the
requested frequency has already been set and return early, resulting in
the correct/new frequency request never making it to the hardware.
Since the operating frequency is liable to this sort of inconsistency,
mark the CPPC driver with CPUFREQ_NEED_UPDATE_LIMITS so that it is always
invoked when a target frequency update is requested.
Recent patches selecting HAVE_RELIABLE_STACKTRACE and HAVE_LIVEPATCH
added them to the end of the ARM64 Kconfig select list. Move them around
to keep this list nearly alphabetically ordered.
Xiang Mei [Thu, 17 Jul 2025 23:01:28 +0000 (16:01 -0700)]
net/sched: sch_qfq: Avoid triggering might_sleep in atomic context in qfq_delete_class
might_sleep could be trigger in the atomic context in qfq_delete_class.
qfq_destroy_class was moved into atomic context locked
by sch_tree_lock to avoid a race condition bug on
qfq_aggregate. However, might_sleep could be triggered by
qfq_destroy_class, which introduced sleeping in atomic context (path:
qfq_destroy_class->qdisc_put->__qdisc_destroy->lockdep_unregister_key
->might_sleep).
Considering the race is on the qfq_aggregate objects, keeping
qfq_rm_from_agg in the lock but moving the left part out can solve
this issue.
Kevin Brodsky [Thu, 19 Jun 2025 16:00:42 +0000 (17:00 +0100)]
arm64: signal: Remove ISB when resetting POR_EL0
POR_EL0 is set to its most permissive value before setting up the
signal frame, to ensure that uaccess succeeds regardless of the
signal stack's pkey.
We are now tolerant to spurious POE faults. This means that we do
not strictly need to issue an ISB after updating POR_EL0, even when
followed by uaccess. The question is whether a fault is likely to
happen or not if the ISB is omitted; in this case the answer seems
to be no. If the regular stack is used, then it should already be
accessible. If the alternate signal stack is used, then a special
(inaccessible) pkey may be used - the assumption is that this
situation is very uncommon.
Remove the ISB to speed up the regular path - this should not have
any functional impact regardless of the scenario.
The mutexes qdev_mutex and chip->mutex are acquired in that order
throughout the driver. To preserve proper lock hierarchy and avoid
potential deadlocks, they must be released in the reverse
order of acquisition.
This change reorders the unlock sequence to first release chip->mutex
followed by qdev_mutex, ensuring consistency with the locking pattern.
[ fixed the code indentations and Fixes tag by tiwai ]
Paolo Abeni [Tue, 22 Jul 2025 09:35:51 +0000 (11:35 +0200)]
Merge branch 'gve-af_xdp-zero-copy-for-dqo-rda'
Joshua Washington says:
====================
gve: AF_XDP zero-copy for DQO RDA
This patch series adds support for AF_XDP zero-copy in the DQO RDA queue
format.
XSK infrastructure is updated to re-post buffers when adding XSK pools
because XSK umem will be posted directly to the NIC, a departure from
the bounce buffer model used in GQI QPL. A registry of XSK pools is
introduced to prevent the usage of XSK pools when in copy mode.
gve: implement DQO RX datapath and control path for AF_XDP zero-copy
Add the RX datapath for AF_XDP zero-copy for DQ RDA. The RX path is
quite similar to that of the normal XDP case. Parallel methods are
introduced to properly handle XSKs instead of normal driver buffers.
To properly support posting from XSKs, queues are destroyed and
recreated, as the driver was initially making use of page pool buffers
instead of the XSK pool memory.
Expose support for AF_XDP zero-copy, as the TX and RX datapaths both
exist.
Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Praveen Kaligineedi <pkaligineedi@google.com> Signed-off-by: Joshua Washington <joshwash@google.com> Signed-off-by: Jeroen de Borst <jeroendb@google.com> Link: https://patch.msgid.link/20250717152839.973004-6-jeroendb@google.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
gve: implement DQO TX datapath for AF_XDP zero-copy
In the descriptor clean path, a number of changes need to be made to
accommodate out of order completions and double completions.
The XSK stack can only handle completions being processed in order, as a
single counter is incremented in xsk_tx_completed to sigify how many XSK
descriptors have been completed. Because completions can come back out
of order in DQ, a separate queue of XSK descriptors must be maintained.
This queue keeps the pending packets in the order that they were written
so that the descriptors can be counted in xsk_tx_completed in the same
order.
For double completions, a new pending packet state and type are
introduced. The new type, GVE_TX_PENDING_PACKET_DQO_XSK, plays an
anlogous role to pre-existing _SKB and _XDP_FRAME pending packet types
for XSK descriptors. The new state, GVE_PACKET_STATE_XSK_COMPLETE,
represents packets for which no more completions are expected. This
includes packets which have received a packet completion or reinjection
completion, as well as packets whose reinjection completion timer have
timed out. At this point, such packets can be counted as part of
xsk_tx_completed() and freed.
Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Praveen Kaligineedi <pkaligineedi@google.com> Signed-off-by: Joshua Washington <joshwash@google.com> Signed-off-by: Jeroen de Borst <jeroendb@google.com> Link: https://patch.msgid.link/20250717152839.973004-5-jeroendb@google.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Relying on xsk_get_pool_from_qid for getting whether zero copy is
enabled on a queue is erroneous, as an XSK pool is registered in
xp_assign_dev whether AF_XDP zero-copy is enabled or not. This becomes
problematic when queues are restarted in copy mode, as all RX queues
with XSKs will register a pool, causing the driver to exercise the
zero-copy codepath.
This patch adds a bitmap to keep track of which queues have zero-copy
enabled.
Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Joshua Washington <joshwash@google.com> Signed-off-by: Jeroen de Borst <jeroendb@google.com> Link: https://patch.msgid.link/20250717152839.973004-4-jeroendb@google.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The existence of both of these xdp_rxq and xsk_rxq is redundant. xdp_rxq
can be used in both the zero-copy mode and the copy mode case. XSK pool
memory model registration is prioritized over normal memory model
registration to ensure that memory model registration happens only once
per queue.
Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Joshua Washington <joshwash@google.com> Signed-off-by: Jeroen de Borst <jeroendb@google.com> Link: https://patch.msgid.link/20250717152839.973004-3-jeroendb@google.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
gve: deduplicate xdp info and xsk pool registration logic
The XDP registration path currently has a lot of reused logic, leading
changes to the codepaths to be unnecessarily complex. gve_reg_xsk_pool
extracts the logic of registering an XSK pool with a queue into a method
that can be used by both XDP_SETUP_XSK_POOL and gve_reg_xdp_info.
gve_unreg_xdp_info is used to undo XDP info registration in the error
path instead of explicitly unregistering the XDP info, as it is more
complete and idempotent.
This patch will be followed by other changes to the XDP registration
logic, and will simplify those changes due to the use of common methods.
Reviewed-by: Willem de Bruijn <willemb@google.com> Signed-off-by: Joshua Washington <joshwash@google.com> Signed-off-by: Jeroen de Borst <jeroendb@google.com> Link: https://patch.msgid.link/20250717152839.973004-2-jeroendb@google.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
selftests/futex: Define SYS_futex on 32-bit architectures with 64-bit time_t
The kernel does not provide sys_futex() on 32-bit architectures that do not
support 32-bit time representations, such as riscv32.
As a result, glibc cannot define SYS_futex, causing compilation failures in
tests that rely on this syscall. Define SYS_futex as SYS_futex_time64 in
such cases to ensure successful compilation and compatibility.
Signed-off-by: Cynthia Huang <cynthia@andestech.com> Signed-off-by: Ben Zong-You Xie <ben717@andestech.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Muhammad Usama Anjum <usama.anjum@collabora.com> Link: https://lore.kernel.org/all/20250710103630.3156130-1-ben717@andestech.com
Miguel Ojeda [Tue, 22 Jul 2025 08:54:59 +0000 (10:54 +0200)]
rust: io: fix broken intra-doc link to missing `flags` module
There is no `mod flags` in this case, unlike others. Instead, they are
associated constants for the `Flags` type.
Thus reword the sentence to fix the broken intra-doc link, providing
an example of constant and linking to it to clarify which ones we are
referring to.
Thomas Huth [Fri, 11 Jul 2025 07:25:53 +0000 (09:25 +0200)]
powerpc: Drop GPL boilerplate text with obsolete FSF address
The FSF does not reside in the Franklin street anymore, so we should not
request the people to write to this address. Fortunately, these header
files already contain a proper SPDX license identifier, so it should be
fine to simply drop all of this license boilerplate code here.
Thomas Weißschuh [Fri, 18 Jul 2025 13:20:25 +0000 (15:20 +0200)]
powerpc: Don't use %pK through printk
In the past %pK was preferable to %p as it would not leak raw pointer
values into the kernel log.
Since commit ad67b74d2469 ("printk: hash addresses printed with %p")
the regular %p has been improved to avoid this issue.
Furthermore, restricted pointers ("%pK") were never meant to be used
through printk(). They can still unintentionally leak raw pointers or
acquire sleeping locks in atomic contexts.
Switch to the regular pointer formatting which is safer and
easier to reason about.
wifi: mac80211: don't require cipher and keylen in gtk rekey
ieee80211_add_gtk_rekey receives a keyconf as an argument, and the
cipher and keylen are taken from there to the new allocated key.
But in rekey, both the cipher and the keylen should be the same as of
the old key, so let ieee80211_add_gtk_rekey find those, so drivers won't
have to fill it in.
wifi: nl80211: Set num_sub_specs before looping through sub_specs
The processing of the struct cfg80211_sar_specs::sub_specs flexible
array requires its counter, num_sub_specs, to be assigned before the
loop in nl80211_set_sar_specs(). Leave the final assignment after the
loop in place in case fewer ended up in the array.
Fixes: aa4ec06c455d ("wifi: cfg80211: use __counted_by where appropriate") Signed-off-by: Kees Cook <kees@kernel.org> Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org> Link: https://patch.msgid.link/20250721183125.work.183-kees@kernel.org Signed-off-by: Johannes Berg <johannes.berg@intel.com>
wifi: mac80211: Write cnt before copying in ieee80211_copy_rnr_beacon()
While I caught the need for setting cnt early in nl80211_parse_rnr_elems()
in the original annotation of struct cfg80211_rnr_elems with __counted_by,
I missed a similar pattern in ieee80211_copy_rnr_beacon(). Fix this by
moving the cnt assignment to before the loop.
Fixes: 7b6d7087031b ("wifi: cfg80211: Annotate struct cfg80211_rnr_elems with __counted_by") Signed-off-by: Kees Cook <kees@kernel.org> Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org> Link: https://patch.msgid.link/20250721182521.work.540-kees@kernel.org Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Mark Brown [Fri, 18 Jul 2025 22:14:52 +0000 (23:14 +0100)]
kselftest/arm64: Handle attempts to disable SM on SME only systems
The ABI for disabling streaming mode via ptrace is to do a write via the
SVE register set. Following the recent round of fixes to the ptrace code
we don't support this operation on systems without SVE, which is detected
as failures by fp-ptrace. Update the program so that it knows that this
operation is not currently supported.
Mark Brown [Fri, 18 Jul 2025 22:14:51 +0000 (23:14 +0100)]
kselftest/arm64: Fix SVE write data generation for SME only systems
fp-ptrace does not handle SME only systems correctly when generating data,
on SME only systems scenarios where we are not in streaming mode will not
have an expected vector length. This leads to attempts to do memcpy()s of
zero byte arrays which can crash, fix this by skipping generation of SVE
data for cases where we do not expect to have an active vector length.
Mark Brown [Fri, 18 Jul 2025 22:14:50 +0000 (23:14 +0100)]
kselftest/arm64: Test SME on SME only systems in fp-ptrace
When checking that the vector extensions are supported fp-ptrace
currently only checks for SVE being supported which means that we get
into a confused half configured state for SME only systems. Check for
SME as well.
Mark Brown [Fri, 18 Jul 2025 22:03:26 +0000 (23:03 +0100)]
kselftest/arm64: Test FPSIMD format data writes via NT_ARM_SVE in fp-ptrace
The NT_ARM_SVE register set supports two data formats, the native SVE one
and an alternative format where we embed a copy of user_fpsimd_data as used
for NT_PRFPREG in the SVE register set. The register data is set as for a
write to NT_PRFPREG and changes in vector length and streaming mode are
handled as for any NT_ARM_SVE write. This has not previously been tested by
fp-ptrace, add coverage of it.
We do not support writes in FPSIMD format for NT_ARM_SSVE so we skip the
test for anything that would leave us in streaming mode.
Mark Brown [Fri, 18 Jul 2025 18:33:16 +0000 (19:33 +0100)]
kselftest/arm64: Allow sve-ptrace to run on SME only systems
Currently the sve-ptrace test program only runs if the system supports
SVE but since SME includes streaming SVE the tests it offers are valid
even on a system that only supports SME. Since the tests already have
individual hwcap checks just remove the top level test and rely on those.
Johannes Berg [Tue, 22 Jul 2025 08:24:52 +0000 (10:24 +0200)]
Merge tag 'ath-next-20250721' of git://git.kernel.org/pub/scm/linux/kernel/git/ath/ath into wireless-next
Jeff Johnson says:
==================
ath.git patches for v6.17
Highlights for some specific drivers include:
ath9k:
Add AHB "of" support
ath11k:
Support device-specific firmware override
Fix potentially reordered access to device memory
ath12k:
Add more Wi-Fi 7 functionality
Add more statistics to DebugFS
Support different memory profiles
Support 802.11 encap/decap offload to firmware
Fix potentially reordered access to device memory
And of course there is the usual set of cleanups and bug fixes across
the entire family of "ath" drivers.
==================
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
When KCOV is enabled all functions get instrumented, unless
the __no_sanitize_coverage attribute is used. To prepare for
__no_sanitize_coverage being applied to __init functions, we have to
handle differences in how GCC's inline optimizations get resolved. For
s390 this exposed a place where the __init annotation was missing but
ended up being "accidentally correct". Fix this cases and force a couple
functions to be inline with __always_inline.
When KCOV is enabled all functions get instrumented, unless
the __no_sanitize_coverage attribute is used. To prepare for
__no_sanitize_coverage being applied to __init functions, we have to
handle differences in how GCC's inline optimizations get resolved. For
arm this exposed several places where __init annotations were missing
but ended up being "accidentally correct". Fix these cases and force
several functions to be inline with __always_inline.
When KCOV is enabled all functions get instrumented, unless
the __no_sanitize_coverage attribute is used. To prepare for
__no_sanitize_coverage being applied to __init functions, we
have to handle differences in how GCC's inline optimizations get
resolved. For mips this requires adding the __init annotation on
init_mips_clocksource().
powerpc/mm/book3s64: Move kfence and debug_pagealloc related calls to __init section
Move a few kfence and debug_pagealloc related functions in hash_utils.c
and radix_pgtable.c to __init sections since these are only invoked once
by an __init function during system initialization.
i.e.
- hash_debug_pagealloc_alloc_slots()
- hash_kfence_alloc_pool()
- hash_kfence_map_pool()
The above 3 functions only gets called by __init htab_initialize().
- alloc_kfence_pool()
- map_kfence_pool()
The above 2 functions only gets called by __init radix_init_pgtable()
To reduce stale data lifetimes, enable CONFIG_INIT_ON_FREE_DEFAULT_ON as
well. This matches the addition of CONFIG_STACKLEAK=y, which is doing
similar for stack memory.
stackleak: Split KSTACK_ERASE_CFLAGS from GCC_PLUGINS_CFLAGS
In preparation for Clang stack depth tracking for KSTACK_ERASE,
split the stackleak-specific cflags out of GCC_PLUGINS_CFLAGS into
KSTACK_ERASE_CFLAGS.
stackleak: Rename stackleak_track_stack to __sanitizer_cov_stack_depth
The Clang stack depth tracking implementation has a fixed name for
the stack depth tracking callback, "__sanitizer_cov_stack_depth", so
rename the GCC plugin function to match since the plugin has no external
dependencies on naming.
In preparation for adding Clang sanitizer coverage stack depth tracking
that can support stack depth callbacks:
- Add the new top-level CONFIG_KSTACK_ERASE option which will be
implemented either with the stackleak GCC plugin, or with the Clang
stack depth callback support.
- Rename CONFIG_GCC_PLUGIN_STACKLEAK as needed to CONFIG_KSTACK_ERASE,
but keep it for anything specific to the GCC plugin itself.
- Rename all exposed "STACKLEAK" names and files to "KSTACK_ERASE" (named
for what it does rather than what it protects against), but leave as
many of the internals alone as possible to avoid even more churn.
While here, also split "prev_lowest_stack" into CONFIG_KSTACK_ERASE_METRICS,
since that's the only place it is referenced from.
Showrya M N [Fri, 27 Jun 2025 11:23:29 +0000 (16:53 +0530)]
scsi: libiscsi: Initialize iscsi_conn->dd_data only if memory is allocated
In case of an ib_fast_reg_mr allocation failure during iSER setup, the
machine hits a panic because iscsi_conn->dd_data is initialized
unconditionally, even when no memory is allocated (dd_size == 0). This
leads invalid pointer dereference during connection teardown.
Fix by setting iscsi_conn->dd_data only if memory is actually allocated.
Signed-off-by: Showrya M N <showrya@chelsio.com> Signed-off-by: Potnuri Bharat Teja <bharat@chelsio.com> Link: https://lore.kernel.org/r/20250627112329.19763-1-showrya@chelsio.com Reviewed-by: Chris Leech <cleech@redhat.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
-Wflex-array-member-not-at-end was introduced in GCC-14, and we are
getting ready to enable it, globally.
We have a flexible struct iwl_tx_cmd_v6 in the middle of a few structs,
but those don't even need the flexible part.
So, we add iwl_tx_cmd_v6_params, that will contain everything except the
flexible array and use this one for the containing structs.
Also, as part of the refactoring remove unused flex array `payload`.
So, with these changes, fix the following warnings:
drivers/net/wireless/intel/iwlwifi/mld/../fw/api/tdls.h:134:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/iwlwifi/mld/../fw/api/tdls.h:53:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/iwlwifi/mld/../fw/api/tx.h:745:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/iwlwifi/mld/../fw/api/tx.h:764:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/iwlwifi/mvm/../fw/api/tdls.h:134:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/iwlwifi/mvm/../fw/api/tdls.h:53:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/iwlwifi/mvm/../fw/api/tx.h:745:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
drivers/net/wireless/intel/iwlwifi/mvm/../fw/api/tx.h:764:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end]
wifi: iwlwifi: fix cmd length when sending WOWLAN_TSC_RSC_PARAM
In iwl_mvm_wowlan_config_rsc_tsc() when calling iwl_mvm_send_cmd_pdu()
we are accidentally passing the size of a pointer rather than the size
of the object pointed by it.
Fix the expression in order to pass the approriate object length.
Fixes: 493681d9f95b ("wifi: iwlwifi: remove support of version 4 of iwl_wowlan_rsc_tsc_params_cmd")
Address-Coverity-ID: 1647627 ("Incorrect expression (SIZEOF_MISMATCH)") Signed-off-by: Antonio Quartulli <antonio@mandelbit.com> Link: https://patch.msgid.link/20250716201911.700-1-antonio@mandelbit.com Signed-off-by: Miri Korenblit <miriam.rachel.korenblit@intel.com>
scsi: scsi_transport_fc: Add comments to describe added 'rport' parameter
Note that there is no executable code altered by this patch.
Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202507181446.aAoFiDm5-lkp@intel.com/ Signed-off-by: Ewan D. Milne <emilne@redhat.com> Link: https://lore.kernel.org/r/20250721164652.335716-1-emilne@redhat.com Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
The following warning traceback is seen if object debugging is enabled
with the new crypto test code.
ODEBUG: object 9000000106237c50 is on stack 9000000106234000, but NOT annotated.
------------[ cut here ]------------
WARNING: lib/debugobjects.c:655 at lookup_object_or_alloc.part.0+0x19c/0x1f4, CPU#0: kunit_try_catch/468
...
This also results in a boot stall when running the code in qemu:loongarch.
Initializing the worker with INIT_WORK_ONSTACK() fixes the problem.
====================
ethtool: rss: support creating and removing contexts via Netlink
This series completes support of RSS configuration via Netlink.
All functionality supported by the IOCTL is now supported by
Netlink. Future series (time allowing) will add:
- hashing on the flow label, which started this whole thing;
- pinning the RSS context to a Netlink socket for auto-cleanup.
The first patch is a leftover held back from previous series
to avoid conflicting with Gal's fix.
Next 4 patches refactor existing code to make reusing it for
context creation possible. 2 patches after that add create
and delete commands. Last but not least the test is extended.
====================
Jakub Kicinski [Thu, 17 Jul 2025 23:43:43 +0000 (16:43 -0700)]
selftests: drv-net: rss_api: context create and delete tests
Add test cases for creating and deleting contexts.
TAP version 13
1..12
ok 1 rss_api.test_rxfh_nl_set_fail
ok 2 rss_api.test_rxfh_nl_set_indir
ok 3 rss_api.test_rxfh_nl_set_indir_ctx
ok 4 rss_api.test_rxfh_indir_ntf
ok 5 rss_api.test_rxfh_indir_ctx_ntf
ok 6 rss_api.test_rxfh_nl_set_key
ok 7 rss_api.test_rxfh_fields
ok 8 rss_api.test_rxfh_fields_set
ok 9 rss_api.test_rxfh_fields_set_xfrm # SKIP no input-xfrm supported
ok 10 rss_api.test_rxfh_fields_ntf
ok 11 rss_api.test_rss_ctx_add
ok 12 rss_api.test_rss_ctx_ntf
# Totals: pass:11 fail:0 xfail:0 xpass:0 skip:1 error:0
projects / thirdparty / kernel / linux.git / log
