summaryrefslogtreecommitdiff
path: root/drivers
AgeCommit message (Collapse)Author
2026-10-15irqchip/mips-gic: Fix recursive acquisition of gic_lock in gic_set_affinity()Benoît Monin
Commit 322a90638768 ("irqchip/mips-gic: Multi-cluster support") made gic_set_affinity() call gic_set_type() when an interrupt's affinity is moved between clusters. gic_set_type() takes gic_lock with raw_spin_lock_irqsave(), but gic_set_affinity() holds gic_lock at that point, so the same raw spinlock is acquired twice, triggering a deadlock. Split gic_set_type() into a gic_set_type_locked() helper that assumes gic_lock is already held. gic_set_type() takes gic_lock and calls the helper, and gic_set_affinity() calls gic_set_type_locked() directly, since it already holds gic_lock. Fixes: 322a90638768 ("irqchip/mips-gic: Multi-cluster support") Signed-off-by: Benoît Monin <benoit.monin@bootlin.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Link: https://patch.msgid.link/20260929-sync-gic-counters-v4-2-ec70c4b60434@bootlin.com
2026-10-15irqchip/mips-gic: Fix unbalanced cm_core_lock in for_each_online_cpu_gic()Benoît Monin
Commit d9e2ed610a60 ("irqchip/mips-gic: Support multi-cluster in for_each_online_cpu_gic()") added a gic_unlock_cluster() call to the macro's loop increment, which unconditionally invokes mips_cm_unlock_other() on multi-cluster systems. However nothing in the loop ever acquires the corresponding mips_cm_lock_other(), so on multi-cluster hardware every invocation of for_each_online_cpu_gic() releases an unheld per-CPU cm_core_lock. With CONFIG_PROVE_LOCKING this triggers a "bad unlock balance detected" warning at boot, e.g. from gic_irq_domain_map() while mapping local interrupts. Only the first occurrence is reported, since the first warning permanently disables lockdep (debug_locks = 0); the unbalanced release itself silently persists. Fix this by moving both the acquire and release into __gic_with_next_online_cpu() so they stay balanced. When advancing to a CPU in a remote cluster, lock the CM redirect block for that cluster via mips_cm_lock_other(); when leaving a remote cluster (or finishing the iteration) release it with mips_cm_unlock_other(). Local-cluster CPUs require no locking, so single-cluster systems are unaffected. This also makes the redirect region behave correctly when accessing local register blocks of CPUs in other clusters. Drop the now-unused gic_unlock_cluster() helper and its call from the for_each_online_cpu_gic() increment. Fixes: d9e2ed610a60 ("irqchip/mips-gic: Support multi-cluster in for_each_online_cpu_gic()") Signed-off-by: Benoît Monin <benoit.monin@bootlin.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Link: https://patch.msgid.link/20260929-sync-gic-counters-v4-1-ec70c4b60434@bootlin.com
2026-10-15irqchip: Add Xilinx Versal NET SMMU CSR interrupt controller driverAnirudha Sarangi
The Xilinx Versal NET SoC implements a Control and Status Register (CSR) block in front of the ARM SMMUv3 to manage EVENTQ, PRIQ and GLOBAL interrupts. Interrupts are enabled, cleared, and acknowledged through this block by writing to the relevant registers. Once an interrupt is acknowledged, it is forwarded to the parent interrupt controller for further processing. Add an irqchip driver for the CSR block. The driver registers as an interrupt controller and chains to the upstream SMMUv3 driver, so that the standard SMMUv3 driver does not require modification while still supporting Xilinx SoCs. The CSR block can also gate the SMMUv3 CMDQ_SYNC completion interrupt (bit 1), but it is intentionally left unsupported. The Linux arm-smmu-v3 driver signals CMDQ_SYNC completion via polling or MSI and never requests a wired CMD_SYNC interrupt, so there is no in-kernel consumer for it. Exposing it as a child line would only risk spurious, never-acked interrupts, and it is omitted from the DT binding for the same reason. This driver is intended to operate in conjunction with the ARM SMMUv3 driver and the kernel's IOMMU DMA infrastructure. These dependencies are not selected directly in Kconfig to avoid introducing dependency recursion. Co-developed-by: Tushar Nimkar <tushar.nimkar@amd.com> Signed-off-by: Anirudha Sarangi <anirudha.sarangi@amd.com> Signed-off-by: Tushar Nimkar <tushar.nimkar@amd.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Link: https://patch.msgid.link/20260928-smmu-changes-v2-v4-2-3c79a14cd806@amd.com
2026-10-15irqchip/jcore-aic: Fix lost IPIs on SMPKarl Mehltretter
Since commit bdf4e2ac295fe ("genirq: Allow per-cpu interrupt sharing for non-overlapping affinities"), handle_percpu_devid_irq() only runs an action whose affinity mask contains the current CPU. Only actions set up through the percpu_devid request functions have such a mask. handle_jcore_irq() passes every per-CPU interrupt to handle_percpu_devid_irq(). The J2 IPI is not a percpu_devid interrupt. j2_prepare_cpus() requests it with request_irq() and IRQF_PERCPU, so its action has no affinity mask and cpumask_test_cpu() reads through a NULL pointer. J2 has no MMU to catch this. The read returns whatever is at address 0, the action is not found, and IPIs are dropped as spurious ("Spurious percpu IRQ20 on CPU1", printed once). Pass only percpu_devid interrupts (the PIT) to handle_percpu_devid_irq() and the other per-CPU interrupts to handle_percpu_irq() again, as before commit d7e3fd658248f ("irqchip/jcore-aic, clocksource/drivers/jcore: Fix jcore-pit interrupt request"). Found with a two-CPU J2 model simulated from the J-Core RTL. Without this change the boot stalls with both CPUs idle shortly after "smp: Brought up 1 node, 2 CPUs". With it, the boot reaches userspace. Fixes: bdf4e2ac295fe ("genirq: Allow per-cpu interrupt sharing for non-overlapping affinities") Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Assisted-by: LLM Tested-by: John Paul Adrian Glaubitz <glaubitz@physik.fu-berlin.de> Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260928035408.25871-1-kmehltretter@gmail.com
2026-10-15irqchip/atmel-aic[5]: Disable interrupts in the irq_pm_shutdown() callbacksKarl Mehltretter
irq_gc_shutdown() calls the irq_pm_shutdown() callbacks from syscore_shutdown() with interrupts enabled. aic_pm_shutdown() and aic5_pm_shutdown() take the generic chip lock with raw_spin_lock(), and irq_gc_eoi() takes the same lock from the interrupt handler. An interrupt from the chip while the lock is held spins in irq_gc_eoi() and hangs the reboot. Lockdep reports this on every reboot of a SAM9X75 Curiosity board: WARNING: inconsistent lock state inconsistent {IN-HARDIRQ-W} -> {HARDIRQ-ON-W} usage. busybox/101 [HC0[0]:SC0[0]:HE1:SE1] takes: c1c2b434 (&gc->lock){?.-.}-{2:2}, at: aic5_pm_shutdown+0x34/0xd8 {IN-HARDIRQ-W} state was registered at: ... irq_gc_eoi+0x28/0x60 handle_fasteoi_irq+0x120/0x214 ... aic5_pm_shutdown from irq_gc_shutdown+0x4c/0x58 irq_gc_shutdown from syscore_shutdown+0x5c/0x9c syscore_shutdown from kernel_restart+0x34/0x78 kernel_restart from sys_reboot+0x180/0x25c Use the irqsave guard, as commit 33517881ede7 ("irqchip/brcmstb-l2: Use _irqsave locking variants in non-interrupt code") did for brcmstb-l2. Fixes: b1479ebb7720 ("irqchip: atmel-aic: Add atmel AIC/AIC5 drivers") Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Assisted-by: LLM Reviewed-by: Radu Rendec <radu@rendec.net> Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260926004124.88221-1-kmehltretter@gmail.com
2026-10-15irqchip/gic: Fix UBSAN shift-out-of-bounds in GIC helpersZhu Ling
When running with UBSAN enabled, enabling a GPIO controller that uses a GIC interrupt as its parent triggers several shift-out-of-bounds warnings: shift-out-of-bounds in drivers/irqchip/irq-gic-common.c:50:21 left shift of 2 by 30 places cannot be represented in type 'int' Similar reports are emitted from gic_poke_irq() and gic_peek_irq() in drivers/irqchip/irq-gic-v3.c. The corresponding GICv2 helpers use the same signed-shift pattern. These masks are generated by shifting signed integer constants, which invokes undefined behavior when bit 31 is selected. Use BIT() to generate the masks with an unsigned type and make their intent explicit. Fixes: f27ecacc54cc ("[ARM] Add support for ARM GIC") Fixes: 5c0c1f08abf0 ("ARM: 6150/1: gic: implement set_type") Fixes: 021f653791ad ("irqchip: gic-v3: Initial support for GICv3") Signed-off-by: Zhu Ling <zhuling2709@phytium.com.cn> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Zenghui Yu (Huawei) <zenghui.yu@linux.dev> Reviewed-by: Radu Rendec <radu@rendec.net> Acked-by: Marc Zyngier <maz@kernel.org> Link: https://lore.kernel.org/r/2fJDwUUYdEf2_eaRa041L9xkT8RkSWFeo7euOnqMbbPhUatLlEaAvGfB6sOspDmXaxO87Eh7bQLV9UolyjbMtZBT1wB2UGypjPOo0Z-RC0Q=@proton.me Link: https://patch.msgid.link/20260929031050.16988-1-zhuling2709@phytium.com.cn
2026-10-15irqchip/gic-v3: Don't register a redistributor that was never countedFuad Tabba
gic_acpi_match_gicc() counts only the enabled GICCs with a non-zero gicr_base_address, and that count sizes redist_regs[], but gic_acpi_parse_madt_gicc() registers every enabled one. On a MADT with no GICR entries, an enabled GICC with a zero GICR base therefore makes gic_acpi_register_redist() write a struct redist_region past the end of the array. Commit fa2dabe57220e ("irqchip/gic-v3: Don't return errors from gic_acpi_match_gicc()") removed the check that kept the two consistent. Its message says such entries are still caught by gic_populate_rdist(), but that runs from gic_cpu_init(), after the write. Skip the entry instead, and mark the CPU's redistributor broken as the online-capable path does, so that gic_check_rdist() rejects the CPU. gic_starting_cpu() runs past cpuhp's last failure point, so without the mark the CPU would come online with its redistributor and CPU interface unconfigured. With an ITS, its_cpu_init() would then dereference a NULL rd_base. Fixes: fa2dabe57220e ("irqchip/gic-v3: Don't return errors from gic_acpi_match_gicc()") Signed-off-by: Fuad Tabba <fuad.tabba@linux.dev> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Marc Zyngier <maz@kernel.org> Link: https://patch.msgid.link/20260925085000.753400-4-fuad.tabba@linux.dev
2026-10-15irqchip/gic-v3-its: Don't free the tables of an enabled ITSFuad Tabba
its_probe_one() enables the ITS before its_init_domain(), the last step that can fail, and the unwind from there frees the tables and the command queue without disabling it. The kernel comes up with an enabled ITS pointing at freed pages. Enable the ITS last. Nothing reaches the new domain before then: this runs from init_IRQ(), and the ITS joins its_nodes only after the enable. Fixes: 4c21f3c26ecc2 ("irqchip: GICv3: ITS: DT probing and initialization") Signed-off-by: Fuad Tabba <fuad.tabba@linux.dev> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Marc Zyngier <maz@kernel.org> Link: https://patch.msgid.link/20260925085000.753400-3-fuad.tabba@linux.dev
2026-10-15irqchip/gic-v3-its: Don't free a vPE table shared with another ITSFuad Tabba
On GICv4.1, its_alloc_tables() copies a sibling's vPE table baser rather than allocating one, but its_free_tables() frees it like the rest. A probe failure on the copying ITS then frees pages the enabled sibling still points GITS_BASER2 at, and the kernel comes up running on that sibling. Mark a shared table so its_free_tables() skips it. Fixes: 5e5168461c22c ("irqchip/gic-v4.1: VPE table (aka GICR_VPROPBASER) allocation") Signed-off-by: Fuad Tabba <fuad.tabba@linux.dev> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Marc Zyngier <maz@kernel.org> Link: https://patch.msgid.link/20260925085000.753400-2-fuad.tabba@linux.dev
2026-10-15irqchip/riscv-imsic: Reject overlapping group index fieldsPengpeng Hou
The IMSIC address layout places the group index above the page offset, guest index and HART index. imsic_parse_fwnode() checks the individual field widths and the upper end of the group field, but does not check that its lower end is above the other fields. Reject a nonzero group field whose shift overlaps the lower fields. Otherwise a firmware-supplied layout can assign the same address bit to different indices when MSI target addresses are constructed. Keep the existing upper-bound check unchanged. The new lower-bound check applies only when group index bits are present. The issue was found by our static-analysis tool. Fixes: 21a8f8a0eb35 ("irqchip: Add RISC-V incoming MSI controller early driver") Signed-off-by: Pengpeng Hou <hppiscas@163.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Assisted-by: gpt 5 Reviewed-by: Anup Patel <anup@brainfault.org> Link: https://patch.msgid.link/caplitmus-rv.v2.imsic.f2c0e5bdc627451c83deb0cf7eb8834a@163.com
2026-10-15irqchip/sifive-plic: Avoid signed shift in interrupt enable maskPengpeng Hou
__plic_toggle() builds a 32-bit interrupt enable mask using a signed integer literal. A valid interrupt selecting bit 31 evaluates 1 << 31, which shifts a signed int into its sign bit. Use BIT(), matching the equivalent mask construction in plic_irq_eoi(). The resulting mask remains representable in u32 on both RV32 and RV64. The issue was found by our static-analysis tool. Fixes: 8237f8bc4f6e ("irqchip: add a SiFive PLIC driver") Signed-off-by: Pengpeng Hou <hppiscas@163.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Assisted-by: gpt 5 Link: https://patch.msgid.link/caplitmus-rv.v2.plic.87a8fc444de44e109889ba28703c76f3@163.com
2026-10-15irqchip/mst-intc: Validate the DT interrupt rangePengpeng Hou
The driver takes an inclusive interrupt range from the device tree without checking its endpoints. A reversed range wraps the unsigned count calculation, and a range larger than MST_INTC_MAX_IRQS exceeds the controller capacity used by the driver. Reject these ranges before calculating nr_irqs or creating the domain. Check the difference after ordering the endpoints so the validation itself cannot wrap. The issue was found by our static-analysis tool. Fixes: ad4c938c92af ("irqchip/irq-mst: Add MStar interrupt controller support") Signed-off-by: Pengpeng Hou <hppiscas@163.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Assisted-by: gpt 5 Reviewed-by: Radu Rendec <radu@rendec.net> Link: https://patch.msgid.link/20260920034236.16229-1-hppiscas@163.com
2026-10-15irqchip: st: Convert to DEFINE_SIMPLE_DEV_PM_OPS()Triet Hoang
Convert the deprecated SIMPLE_DEV_PM_OPS to DEFINE_SIMPLE_DEV_PM_OPS and pm_sleep_ptr(). This allows dropping the __maybe_unused annotations from the PM callbacks, also reduces kernel size in case CONFIG_PM or CONFIG_PM_SLEEP is disabled. Signed-off-by: Triet Hoang <triet.hoang.dev@gmail.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Radu Rendec <radu@rendec.net> Link: https://patch.msgid.link/20260917013835.223139-1-triet.hoang.dev@gmail.com
2026-10-15irqchip/mchp-eic: Fix clock reference leak in mchp_eic_probe()Wentao Liang
of_clk_get_by_name() returns the clock with an elevated reference count that mchp_eic_probe() keeps for the lifetime of the driver, but none of the probe error paths drops it, so the reference is leaked whenever probe fails after the clock has been obtained. Put the reference in the error cleanup. The clk_prepare_enable() failure path must jump there directly as it cannot call clk_disable_unprepare() on a clock that was never enabled. Fixes: 00fa3461c86d ("irqchip/mchp-eic: Add support for the Microchip EIC") Signed-off-by: Wentao Liang <vulab@iscas.ac.cn> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260916190349.2094401-1-vulab@iscas.ac.cn
2026-10-15irqchip/qcom-pdc: Fix linking against QCOM_SCM firmwareArnd Bergmann
Build testing with SCM disabled or in a loadable module fails with a link error: s390-linux-ld: drivers/irqchip/qcom-pdc.o: in function `qcom_pdc_probe': qcom-pdc.c:(.text+0x18b8): undefined reference to `qcom_scm_is_available' qcom-pdc.c:(.text+0x18ce): undefined reference to `qcom_scm_io_writel' As QCOM_SCM can already be enabled for compile testing on non-qualcomm machines, but is always enabled on Qualcomm Arm kernels, use that as the Kconfig dependency and remove the ||COMPILE_TEST check. Fixes: 1b852de05f74 ("irqchip/qcom: Enable compile testing for QCOM_PDC and QCOM_MPM") Fixes: ad01c2b2f291 ("irqchip/qcom-pdc: Configure PDC to pass through mode") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com> Link: https://patch.msgid.link/20260915195948.3521662-1-arnd@kernel.org
2026-10-15irqchip/imx-intmux: Fix runtime PM teardown on removeGuangshuo Li
imx_intmux_probe() enables the IPG clock and then drops its runtime PM reference with pm_runtime_put(). With CONFIG_PM enabled, the runtime suspend callback disables the IPG clock. With CONFIG_PM disabled, the clock remains enabled after probe. imx_intmux_remove() accesses the CHANIER registers without first making sure that the device is runtime active. The device may therefore be runtime suspended with the IPG clock disabled when these registers are accessed. In addition, the remove path only calls pm_runtime_disable() and does not balance the clock enable when CONFIG_PM is disabled. Resume the device and acquire a runtime PM reference before accessing the registers. If resume fails, skip the register accesses but still tear down the chained handlers and IRQ domains. Disable runtime PM afterwards, drop the acquired reference without triggering another runtime suspend, and explicitly disable the IPG clock. This keeps the clock enabled while the hardware registers are accessed and balances the clk_prepare_enable() performed during probe for both CONFIG_PM configurations. This issue was found by manual code inspection. Fixes: bb403111e017 ("irqchip/imx-intmux: Implement intmux runtime power management") Signed-off-by: Guangshuo Li <lgs201920130244@gmail.com> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Frank Li <Frank.Li@nxp.com> Reviewed-by: Radu Rendec <radu@rendec.net> Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260914134211.1734163-1-lgs201920130244@gmail.com
2026-10-15irqchip/sifive-plic: Panic if initialization failsNam Cao
If PLIC initialization fails, the driver unwinds and returns an error. However, that is futile because the system cannot function without the PLIC: all device interrupts are routed through it. Instead of the unnecessary cleanups, panic when PLIC initialization fails. Signed-off-by: Nam Cao <namcao@linutronix.de> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Radu Rendec <radu@rendec.net> Link: https://patch.msgid.link/20260911145037.427896-1-namcao@linutronix.de
2026-10-15irqchip/gic-v3-its: Clear vpt_page after freeing the pending tableFuad Tabba
its_vpe_teardown() frees vpt_page without clearing it, so its vpt_page == NULL guard only holds the first time it runs on a vPE. struct its_vpe outlives a failed KVM_DEV_ARM_VGIC_CTRL_INIT, so a retry that fails again frees the pending table and the vPE ID a second time. Fixes: 325ff3e78c64c ("irqchip/gic-v3-its: Prevent leak in its_vpe_irq_domain_alloc()") Signed-off-by: Fuad Tabba <fuad.tabba@linux.dev> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Marc Zyngier <maz@kernel.org> Reviewed-by: Zenghui Yu (Huawei) <zenghui.yu@linux.dev> Reviewed-by: Radu Rendec <radu@rendec.net> Link: https://patch.msgid.link/20260908142632.2795308-4-fuad.tabba@linux.dev
2026-10-15irqchip/gic-v3-its: Don't clamp nvecs to zero when id_bits is 32Fuad Tabba
GITS_TYPER_IDBITS is five bits, so id_bits reaches 32 and min_t(unsigned int, ...) truncates BIT(32) to 0. nvecs becomes 0, its_lpi_alloc() rejects that, and its_msi_prepare() returns -ENOMEM. On 32-bit ARM the shift is undefined instead. The clamp is correct for every value below the maximum, so only that one value is affected. Fixes: ce9e40a9a5e5c ("irqchip/gic-v3-its: Limit number of per-device MSIs to the range the ITS supports") Signed-off-by: Fuad Tabba <fuad.tabba@linux.dev> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Marc Zyngier <maz@kernel.org> Reviewed-by: Zenghui Yu (Huawei) <zenghui.yu@linux.dev> Reviewed-by: Radu Rendec <radu@rendec.net> Cc: stable@vger.kernel.org Link: https://patch.msgid.link/20260908142632.2795308-3-fuad.tabba@linux.dev
2026-10-15irqchip/gic-v3: Release the partition node in gic_irq_get_fwspec_info()Fuad Tabba
gic_irq_get_fwspec_info() never puts the node it takes from of_find_node_by_phandle(), so every lookup from a four-cell PPI or EPPI specifier naming a partition leaks one. The node is only used for the comparison against gic_data.parts[], and the mask returned points into gic_data, so it can go out of scope with the lookup. Fixes: 68905ea65ceff ("irqchip/gic-v3: Add FW info retrieval support") Signed-off-by: Fuad Tabba <fuad.tabba@linux.dev> Signed-off-by: Thomas Gleixner <tglx@kernel.org> Reviewed-by: Marc Zyngier <maz@kernel.org> Reviewed-by: Zenghui Yu (Huawei) <zenghui.yu@linux.dev> Reviewed-by: Radu Rendec <radu@rendec.net> Link: https://patch.msgid.link/20260908142632.2795308-2-fuad.tabba@linux.dev
2026-10-15net/mlx5: HWS, shrink FTE action array to fitAlexandre Cassen
An FTE keeps the actions it must release in the array used to build them. This is a 1 KB allocation sized for MLX5_FLOW_CONTEXT_ACTION_MAX entries while a rule rarely needs more than three. Keep a copy sized to the actions in use. Signed-off-by: Alexandre Cassen <acassen@corp.free.fr> Reviewed-by: Simon Horman <horms@kernel.org> Reviewed-by: Tariq Toukan <tariqt@nvidia.com> Link: https://patch.msgid.link/20261009052103.3621855-3-acassen@corp.free.fr Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net/mlx5e: TC, free NIC flow parse attributes after offloadAlexandre Cassen
A NIC flow holds a 4 KB allocation for its parse attributes until deletion although nothing reads them after offload. Free them once the rule is in. Signed-off-by: Alexandre Cassen <acassen@corp.free.fr> Reviewed-by: Simon Horman <horms@kernel.org> Reviewed-by: Tariq Toukan <tariqt@nvidia.com> Link: https://patch.msgid.link/20261009052103.3621855-2-acassen@corp.free.fr Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15hwrng: amd - use PMBASE_OFFSET consistentlyThorsten Blum
Use PMBASE_OFFSET instead of the hard-coded 0xF0 and add the missing space to the ioport_map() error message. Signed-off-by: Thorsten Blum <blum@kernel.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2026-10-15hwrng: cctrng - simplify cctrng_probe()Thorsten Blum
Handle the error directly and drop the goto statement. Signed-off-by: Thorsten Blum <blum@kernel.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2026-10-15hwrng: intel - simplify intel_rng_mod_init()Thorsten Blum
Return errors directly to simplify intel_rng_mod_init(). Signed-off-by: Thorsten Blum <blum@kernel.org> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2026-10-15crypto: qat - fix use of destroyed vf2pf_lock on VF restartAhsan Atta
adf_dev_stop_async() sends the RESTARTING_COMPLETE message after adf_dev_down() has returned, but adf_dev_down() already destroyed the lock that message takes. Both run in sequence in the same worker, so every PF restart locks a destroyed mutex. With CONFIG_DEBUG_MUTEXES on a dh895xcc with 32 host-bound VFs this trips on the first PF reset: DEBUG_LOCKS_WARN_ON(lock->magic != lock) WARNING: kernel/locking/mutex.c:159 at mutex_lock+0x171/0x1c0 Workqueue: adf_vf_stop_wq adf_dev_stop_async [intel_qat] Call Trace: adf_gen2_pfvf_send+0x210/0x530 [intel_qat] adf_vf2pf_notify_restart_complete+0x55/0x80 [intel_qat] adf_dev_stop_async+0xdb/0x140 [intel_qat] The notification cannot move before adf_dev_down(): the PF waits for it before its own reset and the VF must tear down its rings first. vf2pf_lock protects the VF2PF CSR and is unrelated to the PF2VF bottom half it is initialised alongside. Tie its lifetime to the accel device, next to state_lock in adf_devmgr_add_dev()/adf_devmgr_rm_dev(), so it stays valid for messages sent outside the device up/down cycle. It is only touched for VFs, which set is_vf before adf_devmgr_add_dev() and keep it set until adf_devmgr_rm_dev(). Fixes: 6f1b5236348f ("crypto: qat - fix recovery flow for VFs") Cc: stable@vger.kernel.org Signed-off-by: Ahsan Atta <ahsan.atta@intel.com> Reviewed-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2026-10-15crypto: qat - initialise state_lock on every path in adf_devmgr_add_dev()Ahsan Atta
state_lock is initialised at the end of adf_devmgr_add_dev(), but the VF on host path whose id mapping already exists returns earlier via 'goto unlock'. That mapping outlives the VF, so this path is taken every time a VF is recreated on a BDF that already hosted one, e.g.: echo 0 > sriov_numvfs echo 32 > sriov_numvfs accel_dev is freshly allocated on each probe, so state_lock is left zeroed and adf_dev_up() takes an uninitialised mutex. With CONFIG_DEBUG_MUTEXES this is reported on a dh895xcc with 32 host-bound VFs: DEBUG_LOCKS_WARN_ON(lock->magic != lock) WARNING: kernel/locking/mutex.c:159 at mutex_lock+0x171/0x1c0 Call Trace: adf_dev_up+0x27/0xe0 [intel_qat] adf_probe+0x700/0xa80 [qat_dh895xccvf] local_pci_probe+0xd5/0x190 The same happens on teardown when adf_dev_down() locks it again. Initialise the lock next to the other per-device state, before the tables are updated, so it is done on every path. Fixes: 1bdc85550a2b ("crypto: qat - fix concurrency issue when device state changes") Cc: stable@vger.kernel.org Signed-off-by: Ahsan Atta <ahsan.atta@intel.com> Reviewed-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2026-10-15hwrng: imx-rngc - Use optional clockJean-Michel Hautbois
Change devm_clk_get() to devm_clk_get_optional() to support platforms where the RNG clock is always enabled and not exposed via the clock framework (such as ColdFire MCF54418). Reviewed-by: Frank Li <Frank.Li@nxp.com> Signed-off-by: Jean-Michel Hautbois <jeanmichel.hautbois@yoseli.org> Reviewed-by: Martin Kaiser <martin@kaiser.cx> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
2026-10-15USB: serial: mxuport: support RS485 mode configurationCrescent Hsieh
Add support for TIOCSRS485 and TIOCGRS485 so that userspace can configure the serial interface mode using struct serial_rs485. Map the serial_rs485 flags to the device interface modes as follows: - RS232 = no flags set - RS422 = SER_RS485_ENABLED | SER_RS485_MODE_RS422 - RS485_2W (half-duplex) = SER_RS485_ENABLED - RS485_4W (full-duplex) = SER_RS485_ENABLED | SER_RS485_RX_DURING_TX Unsupported flags and fields are cleared before applying the configuration. Cache the sanitized per-port configuration and serialize access with the per-port mutex. Signed-off-by: Crescent Hsieh <crescentcy.hsieh@moxa.com> Signed-off-by: Johan Hovold <johan@kernel.org>
2026-10-15USB: serial: mxuport: handle SEND_NEXT transmit flow controlCrescent Hsieh
The device uses a per-port SEND_NEXT request and event to pace host-to-device transmission. During burn-in testing, sustained writes without waiting for this event resulted in per-port buffer overflows and data errors. The issue can be reproduced on existing G1 devices as well. The driver initiates the protocol by setting the SEND_NEXT bit in a bulk-out header and then stops submitting data for that port. The firmware reports a SEND_NEXT event when the corresponding per-port buffer is empty, at which point the driver resumes transmission. Track the payload placed in bulk-out transfers for each port and set the request bit when it reaches bulk_out_size. Use the generic write control callbacks to hold further submissions until the event arrives and to roll back the payload count and hold state if usb_submit_urb() fails. Use bulk_out_size as a common threshold for all device families instead of encoding their firmware buffer sizes in the driver. Signed-off-by: Crescent Hsieh <crescentcy.hsieh@moxa.com> Signed-off-by: Johan Hovold <johan@kernel.org>
2026-10-15USB: serial: add generic write control callbacksCrescent Hsieh
Some USB serial devices use device-specific flow control to determine whether write submission may proceed. The generic write path currently provides no way to pause submission based on driver-specific state. It also provides no way for a driver to restore its state if URB submission fails. Add an optional write_ready() callback to struct usb_serial_driver. Call it with port->lock held after confirming that data is queued and a write URB is available. If the callback returns false, stop write processing and leave the data in the FIFO. The driver must call usb_serial_generic_write_start() when writes may proceed again. Add an optional write_rollback() callback for restoring driver-specific state after usb_submit_urb() fails. Call it with port->lock held after restoring the generic transmit accounting, and pass it the length of the prepared transfer. Drivers which do not provide these callbacks retain the existing generic write behaviour. Signed-off-by: Crescent Hsieh <crescentcy.hsieh@moxa.com> Signed-off-by: Johan Hovold <johan@kernel.org>
2026-10-15IB/IPoIB: Cancel normal flush work on device stopCarolina Jubran
ipoib_ib_dev_flush_normal() can race with ipoib_stop() because the normal flush work does not run under RTNL. Before commit a77a57a1a22a ("IPoIB: Fix deadlock on RTNL in ipoib_stop()"), ipoib_stop() flushed ipoib_workqueue after clearing IPOIB_FLAG_ADMIN_UP and before calling ipoib_ib_dev_down(). This ensured that any in-flight flush work completed before ipoib_ib_dev_down() performed the final clear of IPOIB_FLAG_OPER_UP. After that commit removed the workqueue flush to avoid a deadlock with workqueue tasks taking RTNL, a normal flush can continue running concurrently with ipoib_stop() after IPOIB_FLAG_ADMIN_UP has been cleared. The flush may observe IPOIB_FLAG_ADMIN_UP set and continue running while ipoib_stop() clears IPOIB_FLAG_ADMIN_UP and brings the device down. The flush can then call ipoib_ib_dev_up() and set IPOIB_FLAG_OPER_UP again after the device has already been stopped: flush_normal (ipoib_workqueue) ipoib_stop() (RTNL held) ------------------------------ ------------------------ test_bit(ADMIN_UP) == 1 clear_bit(ADMIN_UP) ipoib_ib_dev_down() clear_bit(OPER_UP) ipoib_ib_dev_up() set_bit(OPER_UP) This leaves the interface with ADMIN_UP clear and OPER_UP set. During device removal, ipoib_ndo_uninit() waits in destroy_workqueue() while holding RTNL, while ipoib_mcast_carrier_on_task() may keep retrying rtnl_trylock() as long as OPER_UP remains set, resulting in a deadlock. Cancel the normal flush work from ipoib_stop() after clearing IPOIB_FLAG_ADMIN_UP and before bringing the device down. A pending normal flush is canceled, while an already running flush is allowed to finish before ipoib_ib_dev_down() performs the final transition to OPER_UP clear. Only the normal flush needs this synchronization. The heavy flush is already serialized with ipoib_stop() by RTNL, while the light flush does not bring the device back up and therefore cannot recreate this race. Fixes: a77a57a1a22a ("IPoIB: Fix deadlock on RTNL in ipoib_stop()") Reported-by: John Popplewell <john.popplewell@nmc2.ai> Signed-off-by: Carolina Jubran <cjubran@nvidia.com> Reviewed-by: Cosmin Ratiu <cratiu@nvidia.com> Signed-off-by: Edward Srouji <edwards@nvidia.com> Link: https://patch.msgid.link/20261014-ipoib-cancel-work-v1-1-6c0faac016b8@nvidia.com Signed-off-by: Leon Romanovsky <leon@kernel.org>
2026-10-15USB: serial: mxuport: add MUX50U-based device supportCrescent Hsieh
Add support for MUX50U-based UPort G2 devices and Platform UARTs. Add their product IDs and list the supported products by family. Encode the number of ports and device family in each USB device ID entry. Use the family to select the firmware image and version offsets, and the method used to determine whether a firmware download is needed. UPort G2 devices use the shared UP firmware image (moxa-up-mux50u.fw), while Platform UARTs use the shared PF firmware image (moxa-pf-mux50u.fw). Platform UARTs are board-level UARTs accessed through the USB bus. The factory firmware on MUX50U-based devices reports its version in a different format from the runtime firmware image, so version comparison cannot reliably determine whether a firmware download is needed. Check the system-register remap state instead to determine whether the device is already running the runtime firmware from SRAM. Keep the existing per-product firmware selection and version comparison for UPort G1 devices. Signed-off-by: Crescent Hsieh <crescentcy.hsieh@moxa.com> Signed-off-by: Johan Hovold <johan@kernel.org>
2026-10-15USB: serial: mxuport: clean up firmware version handlingCrescent Hsieh
Add a small helper for parsing firmware versions and use it for the firmware image. This avoids open-coded firmware image offsets in probe() and retains the existing image-size validation before reading the version bytes. Keep the existing version comparison policy unchanged, but store the version components explicitly so that the version can be printed without unpacking a raw integer. Print the firmware version fields in decimal, as these fields represent readable version components rather than hexadecimal values. Signed-off-by: Crescent Hsieh <crescentcy.hsieh@moxa.com> Signed-off-by: Johan Hovold <johan@kernel.org>
2026-10-15USB: serial: mxuport: clean up product ID definitions and device tableCrescent Hsieh
Rename the product ID definitions after the corresponding product names to make the product mapping explicit. Introduce a helper macro to keep each device-table entry on one line and make the table easier to scan. Signed-off-by: Crescent Hsieh <crescentcy.hsieh@moxa.com> Signed-off-by: Johan Hovold <johan@kernel.org>
2026-10-15microchip: lan865x: move setting of netops to fix crashDavid Yang
lan865x_probe() assigns netdev_ops only after oa_tc6_init(), which connects the internal MAC-PHY through phy_connect_direct(). phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Link: https://patch.msgid.link/20261008022656.3592191-11-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: ti: icssm-prueth: Move setting of netops to fix crashDavid Yang
icssm_prueth_netdev_init() calls of_phy_get_and_connect() before ndev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Link: https://patch.msgid.link/20261008022656.3592191-10-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: ethernet: aeroflex: move setting of netops to fix crashDavid Yang
greth_of_probe() calls greth_mdio_init(), which connects the PHY, before dev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Acked-by: Andreas Larsson <andreas@gaisler.com> Link: https://patch.msgid.link/20261008022656.3592191-9-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: ethoc: move setting of netops to fix crashDavid Yang
ethoc_probe() calls ethoc_mdio_probe(), which connects the PHY, before netdev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Reviewed-by: Daniel Machon <daniel.machon@microchip.com> Link: https://patch.msgid.link/20261008022656.3592191-8-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: au1000: move setting of netops to fix crashDavid Yang
au1000_probe() calls au1000_mii_probe(), which connects the PHY, before dev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Link: https://patch.msgid.link/20261008022656.3592191-7-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: mv643xx: move setting of netops to fix crashDavid Yang
mv643xx_eth_probe() connects the PHY before dev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Daniel Machon <daniel.machon@microchip.com> Link: https://patch.msgid.link/20261008022656.3592191-6-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: ethernet: adi: Move setting of netops to fix crashDavid Yang
adin1140_probe() calls adin1140_phy_init(), which connects the PHY, before netdev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Reviewed-by: Ciprian Regus <ciprian.regus@analog.com> Link: https://patch.msgid.link/20261008022656.3592191-5-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: ethernet: actions: Move setting of netops to fix crashDavid Yang
owl_emac_probe() calls owl_emac_phy_init(), which connects the PHY, before netdev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Link: https://patch.msgid.link/20261008022656.3592191-4-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: hisilicon: hip04_eth: Move setting of netops to fix crashDavid Yang
hip04_mac_probe() calls of_phy_connect() before ndev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Link: https://patch.msgid.link/20261008022656.3592191-3-mmyangfl@gmail.com Reviewed-by: Jijie Shao <shaojijie@huawei.com> Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15net: hisilicon: hisi_femac: Move setting of netops to fix crashDavid Yang
hisi_femac_drv_probe() connects the PHY before dev->netdev_ops is assigned. phy_attach_direct() -> phy_link_topo_add_phy() reads dev->netdev_ops through netdev_need_ops_lock() since the commit in question, so probing crashes with a NULL pointer dereference when CONFIG_NET_SHAPER is enabled. phy_link_topo_add_phy from phy_attach_direct+0xec/0x37c of_phy_get_and_connect from hisi_femac_drv_probe+0x1f4/0x558 Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b Assign netdev_ops before the PHY attach, as was done for emac in commit 7c9f391ec89c ("net: emac: move setting of netops to fix crash"). Fixes: ded86da4bbb7 ("net: ethtool: relax ethnl_req_get_phydev() locking assertion") Signed-off-by: David Yang <mmyangfl@gmail.com> Reviewed-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Reviewed-by: Jijie Shao <shaojijie@huawei.com> Link: https://patch.msgid.link/20261008022656.3592191-2-mmyangfl@gmail.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15accel/ivpu: Load firmware on cold bootAndrzej Kacprowski
Load the firmware image immediately before each cold boot instead of during probe or cold-boot preparation. This moves image loading to a single call site and aligns code for future driver changes. Signed-off-by: Andrzej Kacprowski <andrzej.kacprowski@linux.intel.com> Reviewed-by: Karol Wachowski <karol.wachowski@linux.intel.com> Reviewed-by: Jeff Hugo <jeff.hugo@oss.qualcomm.com> Signed-off-by: Karol Wachowski <karol.wachowski@linux.intel.com> Link: https://patch.msgid.link/20261014101545.3298558-1-andrzej.kacprowski@linux.intel.com
2026-10-15veth: clear rx queue hint in veth_xmitJiayuan Chen
With more tx queues on one end of a veth pair than rx queues on the other, and RPS or generic XDP enabled on the receiving end, we get: veth1 received packet on queue 2, but number of RX queues is 2 WARNING: net/core/dev.c:5212 at get_rps_cpu+0x560/0x1360 Call Trace: <TASK> netif_rx_internal+0x1af/0x4c0 __netif_rx+0x99/0x350 veth_xmit+0x713/0xca0 dev_hard_start_xmit+0x166/0x5f0 __dev_queue_xmit+0x1797/0x42d0 ip_finish_output2+0xa34/0x1f40 __ip_finish_output+0x510/0x7e0 ip_finish_output+0x2f/0x320 ip_output+0x17a/0x3f0 ip_send_skb+0x1bc/0x220 ...... Easy to hit with "ethtool -L veth0 tx 4", "ethtool -L veth1 rx 2", rps_cpus set on veth1 and a few flows sent over the pair [1]. veth_xmit() uses skb->queue_mapping to pick the peer rq, but never clears it before veth_forward_skb(), so the rx side still sees the tx queue index. Everything on the rx side that goes through skb_get_rx_queue(), like get_rps_cpu() and netif_get_rxqueue() for generic XDP, takes it as a recorded rx queue and warns once it is out of range. Clear it before handing the skb to the peer: - It is the tx queue index of this device, it says nothing about the rx queue of the peer. - The two sides don't even agree on the encoding. The tx side stores the index as is, the rx side stores index + 1 so that 0 can mean "not recorded". So tx queue k is read back as rx queue k - 1, and tx queue 0 as "not recorded". - Commit 710ad98c363a ("veth: Do not record rx queue hint in veth_xmit") already decided that veth should not pass any queue hint to the peer. There is no tx->rx queue mapping to preserve, so nothing is lost by clearing it. On NETDEV_TX_BUSY the skb goes back to the qdisc, which looks up the txq from skb->queue_mapping to decide when to retry, so restore it there, next to the existing __skb_push(). [1]: https://lore.kernel.org/netdev/156834bb-8e40-496e-9443-9d515fa18eab@linux.dev/ Fixes: 76ff5cc91935 ("rtnl: allow to specify number of rx and tx queues on device creation") Signed-off-by: Jiayuan Chen <jiayuan.chen@linux.dev> Reviewed-by: Björn Töpel <bjorn@kernel.org> Reviewed-by: Toshiaki Makita <toshiaki.makita1@gmail.com> Reviewed-by: Lorenzo Bianconi <lorenzo.bianconi@oss.qualcomm.com> Link: https://patch.msgid.link/20261009040412.14571-1-jiayuan.chen@linux.dev Signed-off-by: Paolo Abeni <pabeni@redhat.com>
2026-10-15Merge tag 'loongarch-kvm-7.4' of ↵Paolo Bonzini
git://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson into HEAD LoongArch KVM changes for v7.4 1. Add separate vmid support. 2. Simplify MSI interrupt injection. 3. Add crash hypercall support. 4. Allow to set pv_feature until vCPU run. 5. Enable HW page table walk automatically in selftests.
2026-10-15ata: libata-core: Fix NO_FUA quirks shadowed by earlier quirk entriesNiklas Cassel
ata_dev_quirks() uses the first entry in __ata_dev_quirks[] that matches the device and ignores all later entries. Commit 38d431229bfb ("ata: libata: exclude FUA support for known buggy drives") appended three ATA_QUIRK_NO_FUA entries to the end of the table, with model patterns meant to cover whole product lines. Earlier entries already match some or all of these drives, so NO_FUA is not always applied: - "INTEL*SSDSC2CT*" never takes effect. Every model it matches also matches the earlier "INTEL*SSD*" entry, so these drives only get ATA_QUIRK_ZERO_AFTER_TRIM. - "WDC*WD2500J*" does not take effect for "WDC WD2500JD-*" drives, which match the earlier ATA_QUIRK_WD_BROKEN_LPM entry. - "OCZ-VERTEX*" does not take effect for "OCZ-VERTEX" drives with firmware "1.30", which match the earlier ATA_QUIRK_BROKEN_HPA entry. Move the Intel entry in front of "INTEL*SSD*" and keep ATA_QUIRK_ZERO_AFTER_TRIM in it, so these drives stay on the ZERO_AFTER_TRIM whitelist. Add ATA_QUIRK_NO_FUA to the WD2500JD and OCZ Vertex 1.30 entries. The quirks of all other drives are unchanged. libata only uses FUA when it is enabled with libata.fua=1, so systems using the default setting are not affected. Fixes: 38d431229bfb ("ata: libata: exclude FUA support for known buggy drives") Assisted-by: LLM Reviewed-by: Damien Le Moal <dlemoal@kernel.org> Link: https://lore.kernel.org/r/20261013193915.174460-2-cassel@kernel.org Signed-off-by: Niklas Cassel <cassel@kernel.org>
2026-10-15net: axienet: free outstanding TX buffers in axienet_dma_bd_release()Sagi Maimon
axienet_dma_bd_release() walks the RX ring to unmap and free every receive buffer before releasing it, but frees the TX descriptor ring with dma_free_coherent() alone. Any descriptor that axienet_free_tx_chain() had not yet reclaimed still holds its skb and its streaming DMA mapping, and both are lost. axienet_stop() disables TX NAPI and stops the DMA engine before calling it, so nothing reclaims those descriptors afterwards. Bringing the interface down while frames are in flight therefore leaks up to lp->tx_bd_num skbs and mappings each time. axienet_dma_err_handler() already walks the TX ring this way before it restarts the DMA engine: it unmaps every descriptor whose cntrl is still set - axienet_free_tx_chain() clears it on reclaim - frees any skb still attached, and clears the descriptor. Move that loop into a helper, axienet_free_tx_bufs(), and call it from axienet_dma_bd_release() too. This relies on axienet_stop() having stopped the DMA engine first, as the RX walk in the same function already does. The helper frees the skbs with dev_kfree_skb_any(), as drops: in the error handler, which runs from a workqueue, that frees them directly rather than deferring them to softirq as dev_kfree_skb_irq() did. The walk must not run on a ring that is not there. axienet_open() does not check the result of the reset that runs axienet_dma_bd_init(), so when that reset fails tx_bd_v is either still NULL or, after an earlier close, points at the ring that close freed. Clear tx_bd_v and rx_bd_v once their rings are freed, and skip the walk when tx_bd_v is NULL. That also ends the second dma_free_coherent() of a stale ring which the same path already did. On the axienet_dma_bd_init() error path the TX ring has just been allocated zeroed, so the walk does nothing. This was reported by the Sashiko AI review bot. Tested on the AXI Ethernet MAC of an ADVA TimeCard X2 (PCIe card, with the built-in AXI DMA): traffic passes, and after each of ten down/up cycles and five module reloads, all made with traffic running and each running axienet_dma_bd_release(), traffic resumes and nothing is logged. The leak itself was not measured, and neither the failed-reset paths nor axienet_dma_err_handler() were exercised. Fixes: 8a3b7a252dca ("drivers/net/ethernet/xilinx: added Xilinx AXI Ethernet driver") Reviewed-by: Joe Damato <joe@dama.to> Assisted-by: LLM sparse Signed-off-by: Sagi Maimon <maimon.sagi@gmail.com> Reviewed-by: Simon Horman <horms@kernel.org> Link: https://patch.msgid.link/20261007034620.1360542-1-maimon.sagi@gmail.com Signed-off-by: Paolo Abeni <pabeni@redhat.com>