)]}' { "log": [ { "commit": "11e7861d680c3757eab18ec0a474ff680e007dc4", "tree": "a0f68ff072fb1e98fa9593dfb3f25dc23ab9c502", "parents": [ "fb679c832b6497f19fffb8274c419783909c0912", "42a0305ab114975dbad3fe9efea06976dd62d381" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 13:54:26 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 13:54:26 2025 -0700" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm\n\nPull kvm fixes from Paolo Bonzini:\n \"ARM:\n\n - Correctly handle \u0027invariant\u0027 system registers for protected VMs\n\n - Improved handling of VNCR data aborts, including external aborts\n\n - Fixes for handling of FEAT_RAS for NV guests, providing a sane\n fault context during SEA injection and preventing the use of\n RASv1p1 fault injection hardware\n\n - Ensure that page table destruction when a VM is destroyed gives an\n opportunity to reschedule\n\n - Large fix to KVM\u0027s infrastructure for managing guest context loaded\n on the CPU, addressing issues where the output of AT emulation\n doesn\u0027t get reflected to the guest\n\n - Fix AT S12 emulation to actually perform stage-2 translation when\n necessary\n\n - Avoid attempting vLPI irqbypass when GICv4 has been explicitly\n disabled for a VM\n\n - Minor KVM + selftest fixes\n\n RISC-V:\n\n - Fix pte settings within kvm_riscv_gstage_ioremap()\n\n - Fix comments in kvm_riscv_check_vcpu_requests()\n\n - Fix stack overrun when setting vlenb via ONE_REG\n\n x86:\n\n - Use array_index_nospec() to sanitize the target vCPU ID when\n handling PV IPIs and yields as the ID is guest-controlled.\n\n - Drop a superfluous cpumask_empty() check when reclaiming SEV\n memory, as the common case, by far, is that at least one CPU will\n have entered the VM, and wbnoinvd_on_cpus_mask() will naturally\n handle the rare case where the set of have_run_cpus is empty.\n\n Selftests (not KVM):\n\n - Rename the is_signed_type() macro in kselftest_harness.h to\n is_signed_var() to fix a collision with linux/overflow.h. The\n collision generates compiler warnings due to the two macros having\n different meaning\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm: (29 commits)\n KVM: arm64: nv: Fix ATS12 handling of single-stage translation\n KVM: arm64: Remove __vcpu_{read,write}_sys_reg_{from,to}_cpu()\n KVM: arm64: Fix vcpu_{read,write}_sys_reg() accessors\n KVM: arm64: Simplify sysreg access on exception delivery\n KVM: arm64: Check for SYSREGS_ON_CPU before accessing the 32bit state\n RISC-V: KVM: fix stack overrun when loading vlenb\n RISC-V: KVM: Correct kvm_riscv_check_vcpu_requests() comment\n RISC-V: KVM: Fix pte settings within kvm_riscv_gstage_ioremap()\n KVM: arm64: selftests: Sync ID_AA64MMFR3_EL1 in set_id_regs\n KVM: arm64: Get rid of ARM64_FEATURE_MASK()\n KVM: arm64: Make ID_AA64PFR1_EL1.RAS_frac writable\n KVM: arm64: Make ID_AA64PFR0_EL1.RAS writable\n KVM: arm64: Ignore HCR_EL2.FIEN set by L1 guest\u0027s EL2\n KVM: arm64: Handle RASv1p1 registers\n arm64: Add capability denoting FEAT_RASv1p1\n KVM: arm64: Reschedule as needed when destroying the stage-2 page-tables\n KVM: arm64: Split kvm_pgtable_stage2_destroy()\n selftests: harness: Rename is_signed_type() to avoid collision with overflow.h\n KVM: SEV: don\u0027t check have_run_cpus in sev_writeback_caches()\n KVM: arm64: Correctly populate FAR_EL2 on nested SEA injection\n ...\n" }, { "commit": "42a0305ab114975dbad3fe9efea06976dd62d381", "tree": "1a7c17f90272c36b7c922e626a1236fc2ad2eac4", "parents": [ "085e899aa19be2ca512302077346d22736174e6d", "ee372e645178802be7cb35263de941db7b2c5354" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Fri Aug 29 12:57:31 2025 -0400" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Fri Aug 29 12:57:31 2025 -0400" }, "message": "Merge tag \u0027kvmarm-fixes-6.17-1\u0027 of https://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into HEAD\n\nKVM/arm64 changes for 6.17, take #2\n\n - Correctly handle \u0027invariant\u0027 system registers for protected VMs\n\n - Improved handling of VNCR data aborts, including external aborts\n\n - Fixes for handling of FEAT_RAS for NV guests, providing a sane\n fault context during SEA injection and preventing the use of\n RASv1p1 fault injection hardware\n\n - Ensure that page table destruction when a VM is destroyed gives an\n opportunity to reschedule\n\n - Large fix to KVM\u0027s infrastructure for managing guest context loaded\n on the CPU, addressing issues where the output of AT emulation\n doesn\u0027t get reflected to the guest\n\n - Fix AT S12 emulation to actually perform stage-2 translation when\n necessary\n\n - Avoid attempting vLPI irqbypass when GICv4 has been explicitly\n disabled for a VM\n\n - Minor KVM + selftest fixes\n" }, { "commit": "085e899aa19be2ca512302077346d22736174e6d", "tree": "0a1e434bb5c52d2b148172770b5c301b4f564910", "parents": [ "22b2ca023fc41e95afb3c58c16c2418b1ca85d75", "799766208f09f95677a9ab111b93872d414fbad7" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Fri Aug 29 12:57:18 2025 -0400" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Fri Aug 29 12:57:18 2025 -0400" }, "message": "Merge tag \u0027kvm-riscv-fixes-6.17-1\u0027 of https://github.com/kvm-riscv/linux into HEAD\n\nKVM/riscv fixes for 6.17, take #1\n\n- Fix pte settings within kvm_riscv_gstage_ioremap()\n- Fix comments in kvm_riscv_check_vcpu_requests()\n- Fix stack overrun when setting vlenb via ONE_REG\n" }, { "commit": "fb679c832b6497f19fffb8274c419783909c0912", "tree": "9ac278d67ce95ba7fe455e5ed8f053f7d7d2bca8", "parents": [ "2575e638e2d9c70971e5232398ca22afa6f06293", "a6358f8cf64850f3f27857b8ed8c1b08cfc4685c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 09:15:46 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 09:15:46 2025 -0700" }, "message": "Merge tag \u0027efi-fixes-for-v6.17-1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi\n\nPull EFI fixes from Ard Biesheuvel:\n\n - Assorted fixes for the OP-TEE based pseudo-EFI variable store\n\n - Fix for an OOB access when looking up the same non-existing efivarfs\n entry multiple times in parallel\n\n* tag \u0027efi-fixes-for-v6.17-1\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi:\n efivarfs: Fix slab-out-of-bounds in efivarfs_d_compare\n efi: stmm: Drop unneeded null pointer check\n efi: stmm: Drop unused EFI error from setup_mm_hdr arguments\n efi: stmm: Do not return EFI_OUT_OF_RESOURCES on internal errors\n efi: stmm: Fix incorrect buffer allocation method\n" }, { "commit": "2575e638e2d9c70971e5232398ca22afa6f06293", "tree": "b4340fd09187d7017ddfb8c7713f4cea3baba0e8", "parents": [ "469447200aed04c383189b64aa07070be052c48a", "ab529e6ca1f67bcf31f3ea80c72bffde2e9e053e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 08:51:34 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 08:51:34 2025 -0700" }, "message": "Merge tag \u0027v6.17-rc3-smb3-client-fixes\u0027 of git://git.samba.org/sfrench/cifs-2.6\n\nPull smb client fixes from Steve French:\n\n - Fix possible refcount leak in compound operations\n\n - Fix remap_file_range() return code mapping, found by generic/157\n\n* tag \u0027v6.17-rc3-smb3-client-fixes\u0027 of git://git.samba.org/sfrench/cifs-2.6:\n fs/smb: Fix inconsistent refcnt update\n smb3 client: fix return code mapping of remap_file_range\n" }, { "commit": "469447200aed04c383189b64aa07070be052c48a", "tree": "20a29ae38b85d13d17224eaa44b050a3a4f18bc7", "parents": [ "02d6eeedbc36d4b309d5518778071a749ef79c4e", "ae668cd567a6a7622bc813ee0bb61c42bed61ba7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 08:09:34 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 08:09:34 2025 -0700" }, "message": "Merge tag \u0027xfs-fixes-6.17-rc4\u0027 of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux\n\nPull xfs fixes from Carlos Maiolino:\n \"The highlight I\u0027d like to point here is related to the XFS_RT\n Kconfig, which has been updated to be enabled by default now if\n CONFIG_BLK_DEV_ZONED is enabled.\n\n This also contains a few fixes for zoned devices support in XFS,\n specially related to swapon requests in inodes belonging to the zoned\n FS.\n\n A null-ptr dereference fix in the xattr data, due to a mishandling of\n medium errors generated by block devices is also included\"\n\n* tag \u0027xfs-fixes-6.17-rc4\u0027 of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:\n xfs: do not propagate ENODATA disk errors into xattr code\n xfs: reject swapon for inodes on a zoned file system earlier\n xfs: kick off inodegc when failing to reserve zoned blocks\n xfs: remove xfs_last_used_zone\n xfs: Default XFS_RT to Y if CONFIG_BLK_DEV_ZONED is enabled\n" }, { "commit": "02d6eeedbc36d4b309d5518778071a749ef79c4e", "tree": "1014ddfa5b4151a87f9208b430049527c482a6a6", "parents": [ "ec1abfc8333110b9e645cd59eb3bc0541fd5bd2c", "1f3214aae9f49faf495f3836216afbc6c5400b2e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 07:44:14 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 07:44:14 2025 -0700" }, "message": "Merge tag \u0027hid-for-linus-2025082901\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid\n\nPull HID fixes from Jiri Kosina:\n\n - fixes for memory corruption in intel-thc-hid, hid-multitouch,\n hid-mcp2221 and hid-asus (Aaron Ma, Qasim Ijaz, Arnaud Lecomte)\n\n - power management/resume fix for intel-ish-hid (Zhang Lixu)\n\n - driver reinitialization fix for intel-thc-hid (Even Xu)\n\n - ensure that battery level status is reported as soon as possible,\n which is required at least for some Android use-cases (José Expósito)\n\n - quite a few new device ID additions and device-specific quirks\n\n* tag \u0027hid-for-linus-2025082901\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/hid/hid:\n HID: quirks: add support for Legion Go dual dinput modes\n HID: elecom: add support for ELECOM M-DT2DRBK\n HID: logitech: Add ids for G PRO 2 LIGHTSPEED\n HID: input: report battery status changes immediately\n HID: input: rename hidinput_set_battery_charge_status()\n HID: intel-thc-hid: Intel-quicki2c: Enhance driver re-install flow\n HID: hid-ntrig: fix unable to handle page fault in ntrig_report_version()\n HID: asus: fix UAF via HID_CLAIMED_INPUT validation\n hid: fix I2C read buffer overflow in raw_event() for mcp2221\n HID: wacom: Add a new Art Pen 2\n HID: multitouch: fix slab out-of-bounds access in mt_report_fixup()\n HID: Kconfig: Fix spelling mistake \"enthropy\" -\u003e \"entropy\"\n HID: intel-ish-hid: Increase ISHTP resume ack timeout to 300ms\n HID: intel-thc-hid: intel-thc: Fix incorrect pointer arithmetic in I2C regs save\n HID: intel-thc-hid: intel-quicki2c: Fix ACPI dsd ICRS/ISUB length\n" }, { "commit": "ec1abfc8333110b9e645cd59eb3bc0541fd5bd2c", "tree": "acc0a4491eecb77d6376253afaecbf12ea208bc3", "parents": [ "2bd3731542af4d32402fdc0e247153a94efeb682", "ef3e9c91ed87f13dba877a20569f4a0accf0612c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 07:37:21 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 07:37:21 2025 -0700" }, "message": "Merge tag \u0027regulator-fix-v6.17-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator\n\nPull regulator fix from Mark Brown:\n \"One simple fix for the pm8008 driver for poor error handling,\n switching to use a helper which does the right thing in the\n affected case\"\n\n* tag \u0027regulator-fix-v6.17-rc3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator:\n regulator: pm8008: fix probe failure due to negative voltage selector\n" }, { "commit": "2bd3731542af4d32402fdc0e247153a94efeb682", "tree": "73429a3b698afc45468160eb0a1175cc46a2eab9", "parents": [ "18ee2b9b7bd4e2346e467101c973d62300c8ba85", "82b8166171bdebbc74717e4a0cfb4b89cd0510aa" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 07:29:17 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Aug 29 07:29:17 2025 -0700" }, "message": "Merge tag \u0027ata-6.17-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux\n\nPull ata fixes from Damien Le Moal:\n\n - Fix the type of return values to be signed in the ahci_xgen driver\n (Qianfeng)\n\n - Add the mask_port_ext module parameter to the ahci driver.\n\n This is to allow a user to ignore ports that are advertized as\n external (hotplug capable) in favor of lower link power management\n policies instead of the default max_performance for these ports.\n\n This is useful to allow e.g. laptops to go into low power states when\n hooked up to docking station with sata slots, connected with an\n external port for hotplug (me)\n\n* tag \u0027ata-6.17-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/libata/linux:\n ata: ahci_xgene: Use int type for \u0027rc\u0027 to store error codes\n ata: ahci: Allow ignoring the external/hotplug capability of ports\n" }, { "commit": "18ee2b9b7bd4e2346e467101c973d62300c8ba85", "tree": "6cdab460d5094370c1cf0057a7cf4f0359437aa2", "parents": [ "d1cf752d58d59f9222389c14d67951da8e7fbd2b", "42d2abbfa8c40299e047a9b3e5578fdd309dd2ff" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 19:56:32 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 19:56:32 2025 -0700" }, "message": "Merge tag \u0027drm-fixes-2025-08-29\u0027 of https://gitlab.freedesktop.org/drm/kernel\n\nPull drm fixes from Dave Airlie:\n \"Weekly fixes, feels a bit big.\n\n The major piece is msm fixes, then the usual amdgpu/xe along with some\n mediatek and nouveau fixes and a tegra revert.\n\n gpuvm:\n - fix some typos\n\n xe:\n - Fix user-fence race issue\n - Couple xe_vm fixes\n - Don\u0027t trigger rebind on initial dma-buf validation\n - Fix a build issue related to basename() posix vs gnu discrepancy\n\n amdgpu:\n - pin buffers while vmapping\n - UserQ fixes\n - Revert CSA fix\n - SR-IOV fix\n\n nouveau:\n - fix linear modifier\n - remove some dead code\n\n msm:\n - Core/GPU:\n - fix comment doc warning in gpuvm\n - fix build with KMS disabled\n - fix pgtable setup/teardown race\n - global fault counter fix\n - various error path fixes\n - GPU devcoredump snapshot fixes\n - handle in-place VM_BIND remaps to solve turnip vm update race\n - skip re-emitting IBs for unusable VMs\n - Don\u0027t use %pK through printk\n - moved display snapshot init earlier, fixing a crash\n - DPU:\n - Fixed crash in virtual plane checking code\n - Fixed mode comparison in virtual plane checking code\n - DSI:\n - Adjusted width of resulution-related registers\n - Fixed locking issue on 14nm PLLs\n - UBWC (per Bjorn\u0027s ack)\n - Added UBWC configuration for several missing platforms (fixing\n regression)\n\n mediatek:\n - Add error handling for old state CRTC in atomic_disable\n - Fix DSI host and panel bridge pre-enable order\n - Fix device/node reference count leaks in mtk_drm_get_all_drm_priv\n - mtk_hdmi: Fix inverted parameters in some regmap_update_bits calls\n\n tegra:\n - revert dma-buf change\"\n\n* tag \u0027drm-fixes-2025-08-29\u0027 of https://gitlab.freedesktop.org/drm/kernel: (56 commits)\n drm/mediatek: mtk_hdmi: Fix inverted parameters in some regmap_update_bits calls\n drm/amdgpu/userq: fix error handling of invalid doorbell\n drm/amdgpu: update firmware version checks for user queue support\n drm/amd/amdgpu: disable hwmon power1_cap* for gfx 11.0.3 on vf mode\n Revert \"drm/amdgpu: fix incorrect vm flags to map bo\"\n drm/amdgpu/gfx12: set MQD as appriopriate for queue types\n drm/amdgpu/gfx11: set MQD as appriopriate for queue types\n drm/xe: switch to local xbasename() helper\n drm/xe: Don\u0027t trigger rebind on initial dma-buf validation\n drm/xe/vm: Clear the scratch_pt pointer on error\n drm/xe/vm: Don\u0027t pin the vm_resv during validation\n drm/xe/xe_sync: avoid race during ufence signaling\n Revert \"drm/tegra: Use dma_buf from GEM object instance\"\n soc: qcom: use no-UBWC config for MSM8956/76\n soc: qcom: add configuration for MSM8929\n soc: qcom: ubwc: add more missing platforms\n soc: qcom: ubwc: use no-uwbc config for MSM8917\n drm/msm/dpu: Add a null ptr check for dpu_encoder_needs_modeset\n dt-bindings: display/msm: qcom,mdp5: drop lut clock\n drm/gpuvm: fix various typos in .c and .h gpuvm file\n ...\n" }, { "commit": "d1cf752d58d59f9222389c14d67951da8e7fbd2b", "tree": "24803518449f438f5f90de164b4a5c4f9c1bafec", "parents": [ "220374ab2be5a05dc5e35c9a5337698c942916e1", "95a7c5000956f939b86d8b00b8e6b8345f4a9b65" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 18:51:28 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 18:51:28 2025 -0700" }, "message": "Merge tag \u0027block-6.17-20250828\u0027 of git://git.kernel.dk/linux\n\nPull block fixes from Jens Axboe:\n\n - Fix a lockdep spotted issue on recursive locking for zoned writes, in\n case of errors\n\n - Update bcache MAINTAINERS entry address for Coly\n\n - Fix for a ublk release issue, with selftests\n\n - Fix for a regression introduced in this cycle, where it assumed\n q-\u003erq_qos was always set if the bio flag indicated that\n\n - Fix for a regression introduced in this cycle, where loop retrieving\n block device sizes got broken\n\n* tag \u0027block-6.17-20250828\u0027 of git://git.kernel.dk/linux:\n bcache: change maintainer\u0027s email address\n ublk selftests: add --no_ublk_fixed_fd for not using registered ublk char device\n ublk: avoid ublk_io_release() called after ublk char dev is closed\n block: validate QoS before calling __rq_qos_done_bio()\n blk-zoned: Fix a lockdep complaint about recursive locking\n loop: fix zero sized loop for block special file\n" }, { "commit": "220374ab2be5a05dc5e35c9a5337698c942916e1", "tree": "20d3b62cd916fb44b11fb7e9bef24286c861956e", "parents": [ "9c736ace0666efe68efd53fcdfa2c6653c3e0e72", "98b6fa62c84f2e129161e976a5b9b3cb4ccd117b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 18:41:53 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 18:41:53 2025 -0700" }, "message": "Merge tag \u0027io_uring-6.17-20250828\u0027 of git://git.kernel.dk/linux\n\nPull io_uring fixes from Jens Axboe:\n\n - Use the proper type for min_t() in getting the min of the leftover\n bytes and the buffer length.\n\n - As good practice, use READ_ONCE() consistently for reading ring\n provided buffer lengths. Additionally, stop looping for incremental\n commits if a zero sized buffer is hit, as no further progress can be\n made at that point.\n\n* tag \u0027io_uring-6.17-20250828\u0027 of git://git.kernel.dk/linux:\n io_uring/kbuf: always use READ_ONCE() to read ring provided buffer lengths\n io_uring/kbuf: fix signedness in this_len calculation\n" }, { "commit": "9c736ace0666efe68efd53fcdfa2c6653c3e0e72", "tree": "6990a5423824bd785cf9105b1e57a72de86151fe", "parents": [ "fa58e4f6e1a410fc06fcffd84b38b61426509605", "5189446ba995556eaa3755a6e875bc06675b88bd" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 17:35:51 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 17:35:51 2025 -0700" }, "message": "Merge tag \u0027net-6.17-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net\n\nPull networking fixes from Paolo Abeni:\n \"Including fixes from Bluetooth.\n\n Current release - regressions:\n\n - ipv4: fix regression in local-broadcast routes\n\n - vsock: fix error-handling regression introduced in v6.17-rc1\n\n Previous releases - regressions:\n\n - bluetooth:\n - mark connection as closed during suspend disconnect\n - fix set_local_name race condition\n\n - eth:\n - ice: fix NULL pointer dereference on reset\n - mlx5: fix memory leak in hws_pool_buddy_init error path\n - bnxt_en: fix stats context reservation logic\n - hv: fix loss of receive events from host during channel open\n\n Previous releases - always broken:\n\n - page_pool: fix incorrect mp_ops error handling\n\n - sctp: initialize more fields in sctp_v6_from_sk()\n\n - eth:\n - octeontx2-vf: fix max packet length errors\n - idpf: fix Tx flow scheduling to avoid Tx timeouts\n - bnxt_en: fix memory corruption during ifdown\n - ice: fix incorrect counter for buffer allocation failures\n - mlx5: fix lockdep assertion on sync reset unload event\n - fbnic: fixup rtnl_lock and devl_lock handling\n - xgmac: do not enable RX FIFO overflow interrupts\n\n - phy: mscc: fix when PTP clock is register and unregister\n\n Misc:\n\n - add Telit Cinterion LE910C4-WWX new compositions\"\n\n* tag \u0027net-6.17-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (60 commits)\n net: ipv4: fix regression in local-broadcast routes\n net: macb: Disable clocks once\n fbnic: Move phylink resume out of service_task and into open/close\n fbnic: Fixup rtnl_lock and devl_lock handling related to mailbox code\n net: rose: fix a typo in rose_clear_routes()\n l2tp: do not use sock_hold() in pppol2tp_session_get_sock()\n sctp: initialize more fields in sctp_v6_from_sk()\n MAINTAINERS: rmnet: Update email addresses\n net: rose: include node references in rose_neigh refcount\n net: rose: convert \u0027use\u0027 field to refcount_t\n net: rose: split remove and free operations in rose_remove_neigh()\n net: hv_netvsc: fix loss of early receive events from host during channel open.\n net: stmmac: Set CIC bit only for TX queues with COE\n net: stmmac: xgmac: Correct supported speed modes\n net: stmmac: xgmac: Do not enable RX FIFO Overflow interrupts\n net/mlx5e: Set local Xoff after FW update\n net/mlx5e: Update and set Xon/Xoff upon port speed set\n net/mlx5e: Update and set Xon/Xoff upon MTU set\n net/mlx5: Prevent flow steering mode changes in switchdev mode\n net/mlx5: Nack sync reset when SFs are present\n ...\n" }, { "commit": "42d2abbfa8c40299e047a9b3e5578fdd309dd2ff", "tree": "f3e56310c3d4d20c0ac895a3100bd0c6c7381e07", "parents": [ "49862587fa27db253751aadd65a5c9fc1eda1cc7", "c34414883f773412964404d77cd2fea04c6f7d60" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 10:04:26 2025 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 10:05:10 2025 +1000" }, "message": "Merge tag \u0027mediatek-drm-fixes-20250829\u0027 of https://git.kernel.org/pub/scm/linux/kernel/git/chunkuang.hu/linux into drm-fixes\n\nMediatek DRM Fixes - 20250829\n\n1. Add error handling for old state CRTC in atomic_disable\n2. Fix DSI host and panel bridge pre-enable order\n3. Fix device/node reference count leaks in mtk_drm_get_all_drm_priv\n4. mtk_hdmi: Fix inverted parameters in some regmap_update_bits calls\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Chun-Kuang Hu \u003cchunkuang.hu@kernel.org\u003e\nLink: https://lore.kernel.org/r/20250828234116.4960-1-chunkuang.hu@kernel.org\n" }, { "commit": "fa58e4f6e1a410fc06fcffd84b38b61426509605", "tree": "1c6293ac5e67577d5f1dc0322fdbfa8aaf2c568c", "parents": [ "4d28e280981337360b20ac9812bf0eab9224d5a3", "e246518aa24f1460902725e97d0abf574aec6ade" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 16:34:32 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 16:34:32 2025 -0700" }, "message": "Merge tag \u0027pm-6.17-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull power management fix from Rafael Wysocki:\n \"Add missing locking annotations to two recently introduced\n list_for_each_entry_rcu() loops in the core device suspend/resume\n code (Johannes Berg)\"\n\n* tag \u0027pm-6.17-rc4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n PM: sleep: annotate RCU list iterations\n" }, { "commit": "c34414883f773412964404d77cd2fea04c6f7d60", "tree": "e74dbf0974a0746ab416a5d46a15bc4f70cbe96b", "parents": [ "1f403699c40f0806a707a9a6eed3b8904224021a" ], "author": { "name": "Louis-Alexis Eyraud", "email": "louisalexis.eyraud@collabora.com", "time": "Mon Aug 18 16:17:52 2025 +0200" }, "committer": { "name": "Chun-Kuang Hu", "email": "chunkuang.hu@kernel.org", "time": "Thu Aug 28 23:15:41 2025 +0000" }, "message": "drm/mediatek: mtk_hdmi: Fix inverted parameters in some regmap_update_bits calls\n\nIn mtk_hdmi driver, a recent change replaced custom register access\nfunction calls by regmap ones, but two replacements by regmap_update_bits\nwere done incorrectly, because original offset and mask parameters were\ninverted, so fix them.\n\nFixes: d6e25b3590a0 (\"drm/mediatek: hdmi: Use regmap instead of iomem for main registers\")\nSigned-off-by: Louis-Alexis Eyraud \u003clouisalexis.eyraud@collabora.com\u003e\nReviewed-by: CK Hu \u003cck.hu@mediatek.com\u003e\nLink: https://patchwork.kernel.org/project/dri-devel/patch/20250818-mt8173-fix-hdmi-issue-v1-1-55aff9b0295d@collabora.com/\nSigned-off-by: Chun-Kuang Hu \u003cchunkuang.hu@kernel.org\u003e\n" }, { "commit": "49862587fa27db253751aadd65a5c9fc1eda1cc7", "tree": "8b886575e70a71fd95abbdbb7b51b11b7cd0c86c", "parents": [ "4b1c24ef50bc1d4f0b17d85749809d61cbe45be2", "3cf6147f2b51a569761e1ef010efbd891e3a3a15" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 09:05:16 2025 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 09:05:18 2025 +1000" }, "message": "Merge tag \u0027drm-msm-fixes-2025-08-26\u0027 of https://gitlab.freedesktop.org/drm/msm into drm-fixes\n\nFixes for v6.17-rc4\n\nCore/GPU:\n- fix comment doc warning in gpuvm\n- fix build with KMS disabled\n- fix pgtable setup/teardown race\n- global fault counter fix\n- various error path fixes\n- GPU devcoredump snapshot fixes\n- handle in-place VM_BIND remaps to solve turnip vm update race\n- skip re-emitting IBs for unusable VMs\n- Don\u0027t use %pK through printk\n- moved display snapshot init earlier, fixing a crash\n\nDPU:\n- Fixed crash in virtual plane checking code\n- Fixed mode comparison in virtual plane checking code\n\nDSI:\n- Adjusted width of resulution-related registers\n- Fixed locking issue on 14nm PLLs\n\nUBWC (per Bjorn\u0027s ack)\n- Added UBWC configuration for several missing platforms (fixing\n regression)\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\nFrom: Rob Clark \u003crob.clark@oss.qualcomm.com\u003e\nLink: https://lore.kernel.org/r/CACSVV02+u1VW1dzuz6JWwVEfpgTj6Y-JXMH+vX43KsKTVsW+Yg@mail.gmail.com\n" }, { "commit": "4d28e280981337360b20ac9812bf0eab9224d5a3", "tree": "6b7356ffb2d3e83a26e7b1177853ad84f33429df", "parents": [ "5b9f3b013b33e73c6f8f2bbdc6b3fd87745b40af", "89a2d212bdb4bc29bed8e7077abe054b801137ea" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 16:04:14 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 16:04:14 2025 -0700" }, "message": "Merge tag \u0027dma-mapping-6.17-2025-08-28\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mszyprowski/linux\n\nPull dma-mapping fixes from Marek Szyprowski:\n\n - another small fix for arm64 systems with memory encryption (Shanker\n Donthineni)\n\n - fix for arm32 systems with non-standard CMA configuration (Oreoluwa\n Babatunde)\n\n* tag \u0027dma-mapping-6.17-2025-08-28\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mszyprowski/linux:\n dma/pool: Ensure DMA_DIRECT_REMAP allocations are decrypted\n of: reserved_mem: Restructure call site for dma_contiguous_early_fixup()\n" }, { "commit": "4b1c24ef50bc1d4f0b17d85749809d61cbe45be2", "tree": "0c1d876502825e86f6812e87f57c8f04ea67a847", "parents": [ "60d98e1a8dec15f6b9b3fb4174645d1e48a64c79", "c767d74a9cdd1042046d02319d16b85d9aa8a8aa" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 08:50:27 2025 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 08:50:44 2025 +1000" }, "message": "Merge tag \u0027amd-drm-fixes-6.17-2025-08-28\u0027 of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes\n\namd-drm-fixes-6.17-2025-08-28:\n\namdgpu:\n- UserQ fixes\n- Revert CSA fix\n- SR-IOV fix\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Alex Deucher \u003calexander.deucher@amd.com\u003e\nLink: https://lore.kernel.org/r/20250828173904.75850-1-alexander.deucher@amd.com\n" }, { "commit": "5b9f3b013b33e73c6f8f2bbdc6b3fd87745b40af", "tree": "e8133077b8a32077e4920832a7ea76de54848f63", "parents": [ "606c2cf67b6b8421ec40769f03d8b30759df27c2", "b3dcc9d1d806fb1e175f85978713eef868531da4" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 15:46:06 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 15:46:06 2025 -0700" }, "message": "Merge tag \u0027fixes-2025-08-28\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock\n\nPull memblock fixes from Mike Rapoport:\n\n - printk cleanups in memblock and numa_memblks\n\n - update kernel-doc for MEMBLOCK_RSRV_NOINIT to be more accurate and\n detailed\n\n* tag \u0027fixes-2025-08-28\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock:\n memblock: fix kernel-doc for MEMBLOCK_RSRV_NOINIT\n mm: numa,memblock: Use SZ_1M macro to denote bytes to MB conversion\n mm/numa_memblks: Use pr_debug instead of printk(KERN_DEBUG)\n" }, { "commit": "60d98e1a8dec15f6b9b3fb4174645d1e48a64c79", "tree": "036a9e16307bb8651dd051e6feb83273a30b7cdf", "parents": [ "59b00024ce5baeecb3a7becf11271bbee8a85f62", "16fdb3cc6af8460f23a706512c6f5e7dfdd4f338" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 08:44:10 2025 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 08:44:53 2025 +1000" }, "message": "Merge tag \u0027drm-misc-fixes-2025-08-28\u0027 of https://gitlab.freedesktop.org/drm/misc/kernel into drm-fixes\n\nSeveral nouveau fixes to remove unused code, fix an error path and be\nless restrictive with the formats it accepts. A fix for amdgpu to pin\nvmapped dma-buf, and a revert for tegra for a regression in the dma-buf\n/ GEM code.\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Maxime Ripard \u003cmripard@redhat.com\u003e\nLink: https://lore.kernel.org/r/20250828-hypersonic-colorful-squirrel-64f04b@houat\n" }, { "commit": "606c2cf67b6b8421ec40769f03d8b30759df27c2", "tree": "86c99ea7f32dafea780992275aa8cafe2e7e9eef", "parents": [ "ebf2bfec412ad293a0b118fb1a20a551088ebc9b", "8763d2257f5231cfdfd8a53594647927dbf8bb06" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 15:39:06 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 15:39:06 2025 -0700" }, "message": "Merge tag \u0027powerpc-6.17-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux\n\nPull powerpc fixes from Madhavan Srinivasan:\n\n - Merge two CONFIG_POWERPC64_CPU entries in Kconfig.cputype\n\n - Replace extra-y to always-y in Makefile\n\n - Cleanup to use dev_fwnode helper\n\n - Fix misleading comment in kvmppc_prepare_to_enter()\n\n - misc cleanup and fixes\n\nThanks to Amit Machhiwal, Andrew Donnellan, Christophe Leroy, Gautam\nMenghani, Jiri Slaby (SUSE), Masahiro Yamada, Shrikanth Hegde, Stephen\nRothwell, Venkat Rao Bagalkote, and Xichao Zhao\n\n* tag \u0027powerpc-6.17-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:\n powerpc/boot/install.sh: Fix shellcheck warnings\n powerpc/prom_init: Fix shellcheck warnings\n powerpc/kvm: Fix ifdef to remove build warning\n powerpc: unify two CONFIG_POWERPC64_CPU entries in the same choice block\n powerpc: use always-y instead of extra-y in Makefiles\n powerpc/64: Drop unnecessary \u0027rc\u0027 variable\n powerpc: Use dev_fwnode()\n KVM: PPC: Fix misleading interrupts comment in kvmppc_prepare_to_enter()\n" }, { "commit": "ebf2bfec412ad293a0b118fb1a20a551088ebc9b", "tree": "39def63336071fc85e8941400ce8c855cd5d43d6", "parents": [ "07d9df80082b8d1f37e05658371b087cb6738770" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 15:16:16 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Aug 28 15:16:16 2025 -0700" }, "message": "MAINTAINERS: mark bcachefs externally maintained\n\nAs per many long discussion threads, public and private.\n\nSigned-off-by: Linus Torvalds \u003ctorbalds@linux-foundation.org\u003e\n" }, { "commit": "59b00024ce5baeecb3a7becf11271bbee8a85f62", "tree": "0f4167d38d72cd20289878118b165ca2d4a28ec5", "parents": [ "1b237f190eb3d36f52dffe07a40b5eb210280e00", "75671d90fde8c78e940e15a1366a50ece56c6b69" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 07:06:31 2025 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Aug 29 07:08:11 2025 +1000" }, "message": "Merge tag \u0027drm-xe-fixes-2025-08-27\u0027 of https://gitlab.freedesktop.org/drm/xe/kernel into drm-fixes\n\n- Fix user-fence race issue (Zbigniew)\n- Couple xe_vm fixes (Thomas)\n- Don\u0027t trigger rebind on initial dma-buf validation (Brost)\n- Fix a build issue related to basename() posix vs gnu discrepancy (Carlos)\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n\nFrom: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nLink: https://lore.kernel.org/r/aK8oalcIU-zQOfws@intel.com\n" }, { "commit": "ee372e645178802be7cb35263de941db7b2c5354", "tree": "889aa130d973dc3619f05c4cc2dead3ec72fde75", "parents": [ "3328d17e70545f83515b07851302d48c85c466b0" ], "author": { "name": "Marc Zyngier", "email": "maz@kernel.org", "time": "Sat Aug 09 15:48:10 2025 +0100" }, "committer": { "name": "Oliver Upton", "email": "oliver.upton@linux.dev", "time": "Thu Aug 28 12:44:42 2025 -0700" }, "message": "KVM: arm64: nv: Fix ATS12 handling of single-stage translation\n\nVolodymyr reports that using a Xen DomU as a nested guest (where\nHCR_EL2.E2H \u003d\u003d 0), ATS12 results in a translation that stops at\nthe L2\u0027s S1, which isn\u0027t something you\u0027d normally expects.\n\nComparing the code against the spec proves to be illuminating,\nand suggests that the author of such code must have been tired,\ncross-eyed, drunk, or maybe all of the above.\n\nThe gist of it is that, apart from HCR_EL2.VM or HCR_EL2.DC being\n0, only the use of the EL2\u00260 translation regime limits the walk\nto S1 only, and that we must finish the S2 walk in any other case.\nWhich solves the above issue, as E2H\u003d\u003d0 indicates that ATS12 walks\nthe EL1\u00260 translation regime.\n\nExplicitly checking for EL2\u00260 fixes this.\n\nReported-by: Volodymyr Babchuk \u003cvolodymyr_babchuk@epam.com\u003e\nSuggested-by: Oliver Upton \u003coliver.upton@linux.dev\u003e\nSigned-off-by: Marc Zyngier \u003cmaz@kernel.org\u003e\nFixes: be04cebf3e788 (\"KVM: arm64: nv: Add emulation of AT S12E{0,1}{R,W}\")\nLink: https://lore.kernel.org/r/20250806141707.3479194-2-volodymyr_babchuk@epam.com\nLink: https://lore.kernel.org/r/20250809144811.2314038-2-maz@kernel.org\nSigned-off-by: Oliver Upton \u003coliver.upton@linux.dev\u003e\n" }, { "commit": "3328d17e70545f83515b07851302d48c85c466b0", "tree": "de34b368e52d4dee06ae72c1cb92e9ae18370bcb", "parents": [ "ec0ab059d4359b2dad69f4ef90fa58a0bcdef525" ], "author": { "name": "Marc Zyngier", "email": "maz@kernel.org", "time": "Sun Aug 17 13:19:26 2025 +0100" }, "committer": { "name": "Oliver Upton", "email": "oliver.upton@linux.dev", "time": "Thu Aug 28 11:39:48 2025 -0700" }, "message": "KVM: arm64: Remove __vcpu_{read,write}_sys_reg_{from,to}_cpu()\n\nThere is no point having __vcpu_{read,write}_sys_reg_{from,to}_cpu()\nexposed to the rest of the kernel, as the only callers are in\nsys_regs.c.\n\nMove them where they below, which is another opportunity to\nsimplify things a bit.\n\nSigned-off-by: Marc Zyngier \u003cmaz@kernel.org\u003e\nLink: https://lore.kernel.org/r/20250817121926.217900-5-maz@kernel.org\nSigned-off-by: Oliver Upton \u003coliver.upton@linux.dev\u003e\n" }, { "commit": "ec0ab059d4359b2dad69f4ef90fa58a0bcdef525", "tree": "2d8e2780ca01732033ccc88f15aff21f5c4bf5c2", "parents": [ "e3f6836a632e6d68201b6bc6e02bda92b00f3f57" ], "author": { "name": "Marc Zyngier", "email": "maz@kernel.org", "time": "Sun Aug 17 13:19:25 2025 +0100" }, "committer": { "name": "Oliver Upton", "email": "oliver.upton@linux.dev", "time": "Thu Aug 28 11:39:48 2025 -0700" }, "message": "KVM: arm64: Fix vcpu_{read,write}_sys_reg() accessors\n\nVolodymyr reports (again!) that under some circumstances (E2H\u003d\u003d0,\nwalking S1 PTs), PAR_EL1 doesn\u0027t report the value of the latest\nwalk in the CPU register, but that instead the value is written to\nthe backing store.\n\nFurther investigation indicates that the root cause of this is\nthat a group of registers (PAR_EL1, TPIDR*_EL{0,1}, the *32_EL2 dregs)\nshould always be considered as \"on CPU\", as they are not remapped\nbetween EL1 and EL2.\n\nWe fail to treat them accordingly, and end-up considering that\nthe register (PAR_EL1 in this example) should be written to memory\ninstead of in the register.\n\nWhile it would be possible to quickly work around it, it is obvious\nthat the way we track these things at the moment is pretty horrible,\nand could do with some improvement.\n\nRevamp the whole thing by:\n\n- defining a location for a register (memory, cpu), potentially\n depending on the state of the vcpu\n\n- define a transformation for this register (mapped register, potential\n translation, special register needing some particular attention)\n\n- convey this information in a structure that can be easily passed\n around\n\nAs a result, the accessors themselves become much simpler, as the\nstate is explicit instead of being driven by hard-to-understand\nconventions.\n\nWe get rid of the \"pure EL2 register\" notion, which wasn\u0027t very\nuseful, and add sanitisation of the values by applying the RESx\nmasks as required, something that was missing so far.\n\nAnd of course, we add the missing registers to the list, with the\nindication that they are always loaded.\n\nReported-by: Volodymyr Babchuk \u003cvolodymyr_babchuk@epam.com\u003e\nSigned-off-by: Marc Zyngier \u003cmaz@kernel.org\u003e\nFixes: fedc612314acf (\"KVM: arm64: nv: Handle virtual EL2 registers in vcpu_read/write_sys_reg()\")\nLink: https://lore.kernel.org/r/20250806141707.3479194-3-volodymyr_babchuk@epam.com\nLink: https://lore.kernel.org/r/20250817121926.217900-4-maz@kernel.org\nSigned-off-by: Oliver Upton \u003coliver.upton@linux.dev\u003e\n" }, { "commit": "e3f6836a632e6d68201b6bc6e02bda92b00f3f57", "tree": "ca578e6eba707b831b001e9b479280159169e8af", "parents": [ "b720269334f55f2e683fbb8f1f065a12f2c72eb6" ], "author": { "name": "Marc Zyngier", "email": "maz@kernel.org", "time": "Sun Aug 17 13:19:24 2025 +0100" }, "committer": { "name": "Oliver Upton", "email": "oliver.upton@linux.dev", "time": "Thu Aug 28 11:39:48 2025 -0700" }, "message": "KVM: arm64: Simplify sysreg access on exception delivery\n\nDistinguishing between NV and VHE is slightly pointless, and only\nserves as an extra complication, or a way to introduce bugs, such\nas the way SPSR_EL1 gets written without checking for the state\nbeing resident.\n\nGet rid if this silly distinction, and fix the bug in one go.\n\nSigned-off-by: Marc Zyngier \u003cmaz@kernel.org\u003e\nLink: https://lore.kernel.org/r/20250817121926.217900-3-maz@kernel.org\nSigned-off-by: Oliver Upton \u003coliver.upton@linux.dev\u003e\n" }, { "commit": "b720269334f55f2e683fbb8f1f065a12f2c72eb6", "tree": "d7ece2fb0f4abd0a9ee6a1913120ef909efc7742", "parents": [ "01860bcc53432d8b9b92a72939b35679ac24059f" ], "author": { "name": "Marc Zyngier", "email": "maz@kernel.org", "time": "Sun Aug 17 13:19:23 2025 +0100" }, "committer": { "name": "Oliver Upton", "email": "oliver.upton@linux.dev", "time": "Thu Aug 28 11:39:48 2025 -0700" }, "message": "KVM: arm64: Check for SYSREGS_ON_CPU before accessing the 32bit state\n\nJust like c6e35dff58d3 (\"KVM: arm64: Check for SYSREGS_ON_CPU before\naccessing the CPU state\") fixed the 64bit state access, add a check\nfor the 32bit state actually being on the CPU before writing it.\n\nSigned-off-by: Marc Zyngier \u003cmaz@kernel.org\u003e\nLink: https://lore.kernel.org/r/20250817121926.217900-2-maz@kernel.org\nSigned-off-by: Oliver Upton \u003coliver.upton@linux.dev\u003e\n" }, { "commit": "95a7c5000956f939b86d8b00b8e6b8345f4a9b65", "tree": "58aee4fe007c2c38a991993fb72cb99e5dd5d482", "parents": [ "9b2785ea8592f239836405de023c75c4f3f5ce00" ], "author": { "name": "Coly Li", "email": "colyli@suse.de", "time": "Thu Aug 28 23:48:35 2025 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Aug 28 10:05:37 2025 -0600" }, "message": "bcache: change maintainer\u0027s email address\n\nChange to my new email address on fnnas.com.\n\nSigned-off-by: Coly Li \u003ccolyli@fnnas.com\u003e\nLink: https://lore.kernel.org/r/20250828154835.32926-1-colyli@kernel.org\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "9b2785ea8592f239836405de023c75c4f3f5ce00", "tree": "6b090b91045ba0556d4cfefbfdf909ed8353b3de", "parents": [ "c5c5eb24ed6177fc0ef4bb75fc18d07a99c1d3f0" ], "author": { "name": "Ming Lei", "email": "ming.lei@redhat.com", "time": "Wed Aug 27 20:16:00 2025 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Aug 28 07:56:57 2025 -0600" }, "message": "ublk selftests: add --no_ublk_fixed_fd for not using registered ublk char device\n\nAdd a new command line option --no_ublk_fixed_fd that excludes the ublk\ncontrol device (/dev/ublkcN) from io_uring\u0027s registered files array.\nWhen this option is used, only backing files are registered starting\nfrom index 1, while the ublk control device is accessed using its raw\nfile descriptor.\n\nAdd ublk_get_registered_fd() helper function that returns the appropriate\nfile descriptor for use with io_uring operations.\n\nKey optimizations implemented:\n- Cache UBLKS_Q_NO_UBLK_FIXED_FD flag in ublk_queue.flags to avoid\n reading dev-\u003eno_ublk_fixed_fd in fast path\n- Cache ublk char device fd in ublk_queue.ublk_fd for fast access\n- Update ublk_get_registered_fd() to use ublk_queue * parameter\n- Update io_uring_prep_buf_register/unregister() to use ublk_queue *\n- Replace ublk_device * access with ublk_queue * access in fast paths\n\nAlso pass --no_ublk_fixed_fd to test_stress_04.sh for covering\nplain ublk char device mode.\n\nSigned-off-by: Ming Lei \u003cming.lei@redhat.com\u003e\nLink: https://lore.kernel.org/r/20250827121602.2619736-3-ming.lei@redhat.com\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "c5c5eb24ed6177fc0ef4bb75fc18d07a99c1d3f0", "tree": "4d6cbf5e10f956dcbbfa8d59a9c4fcd7cc178b59", "parents": [ "e3ef9445cd9d90e43de0bd3cd55d437773dfd139" ], "author": { "name": "Ming Lei", "email": "ming.lei@redhat.com", "time": "Wed Aug 27 20:15:59 2025 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Aug 28 07:56:57 2025 -0600" }, "message": "ublk: avoid ublk_io_release() called after ublk char dev is closed\n\nWhen running test_stress_04.sh, the following warning is triggered:\n\nWARNING: CPU: 1 PID: 135 at drivers/block/ublk_drv.c:1933 ublk_ch_release+0x423/0x4b0 [ublk_drv]\n\nThis happens when the daemon is abruptly killed:\n\n- some references may still be held, because registering IO buffer\ndoesn\u0027t grab ublk char device reference\n\nOR\n\n- io-\u003etask_registered_buffers won\u0027t be cleared because io buffer is\nreleased from non-daemon context\n\nFor zero-copy and auto buffer register modes, I/O reference crosses\nsyscalls, so IO reference may not be dropped naturally when ublk server is\nkilled abruptly. However, when releasing io_uring context, it is guaranteed\nthat the reference is dropped finally, see io_sqe_buffers_unregister() from\nio_ring_ctx_free().\n\nFix this by adding ublk_drain_io_references() that:\n- Waits for active I/O references dropped in async way by scheduling\n work function, for avoiding ublk dev and io_uring file\u0027s release\n dependency\n- Reinitializes io-\u003eref and io-\u003etask_registered_buffers to clean state\n\nThis ensures the reference count state is clean when ublk_queue_reinit()\nis called, preventing the warning and potential use-after-free.\n\nFixes: 1f6540e2aabb (\"ublk: zc register/unregister bvec\")\nFixes: 1ceeedb59749 (\"ublk: optimize UBLK_IO_UNREGISTER_IO_BUF on daemon task\")\nFixes: 8a8fe42d765b (\"ublk: optimize UBLK_IO_REGISTER_IO_BUF on daemon task\")\nSigned-off-by: Ming Lei \u003cming.lei@redhat.com\u003e\nLink: https://lore.kernel.org/r/20250827121602.2619736-2-ming.lei@redhat.com\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "98b6fa62c84f2e129161e976a5b9b3cb4ccd117b", "tree": "7a4d2a10e89a2c81bcfe9e3d8686bb0e8adbef5a", "parents": [ "c64eff368ac676e8540344d27a3de47e0ad90d21" ], "author": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Aug 27 15:27:30 2025 -0600" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Thu Aug 28 05:48:34 2025 -0600" }, "message": "io_uring/kbuf: always use READ_ONCE() to read ring provided buffer lengths\n\nSince the buffers are mapped from userspace, it is prudent to use\nREAD_ONCE() to read the value into a local variable, and use that for\nany other actions taken. Having a stable read of the buffer length\navoids worrying about it changing after checking, or being read multiple\ntimes.\n\nSimilarly, the buffer may well change in between it being picked and\nbeing committed. Ensure the looping for incremental ring buffer commit\nstops if it hits a zero sized buffer, as no further progress can be made\nat that point.\n\nFixes: ae98dbf43d75 (\"io_uring/kbuf: add support for incremental buffer consumption\")\nLink: https://lore.kernel.org/io-uring/tencent_000C02641F6250C856D0C26228DE29A3D30A@qq.com/\nReported-by: Qingyue Zhang \u003cchunzhennn@qq.com\u003e\nReported-by: Suoxing Zhang \u003caftern00n@qq.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "5189446ba995556eaa3755a6e875bc06675b88bd", "tree": "b49ec45aeafb8c87e896c2f843eb589959a2ebb3", "parents": [ "dac978e51cce0c1f00a14c4a82f81d387f79b2d4" ], "author": { "name": "Oscar Maes", "email": "oscmaes92@gmail.com", "time": "Wed Aug 27 08:23:21 2025 +0200" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Aug 28 10:52:30 2025 +0200" }, "message": "net: ipv4: fix regression in local-broadcast routes\n\nCommit 9e30ecf23b1b (\"net: ipv4: fix incorrect MTU in broadcast routes\")\nintroduced a regression where local-broadcast packets would have their\ngateway set in __mkroute_output, which was caused by fi \u003d NULL being\nremoved.\n\nFix this by resetting the fib_info for local-broadcast packets. This\npreserves the intended changes for directed-broadcast packets.\n\nCc: stable@vger.kernel.org\nFixes: 9e30ecf23b1b (\"net: ipv4: fix incorrect MTU in broadcast routes\")\nReported-by: Brett A C Sheffield \u003cbacs@librecast.net\u003e\nCloses: https://lore.kernel.org/regressions/20250822165231.4353-4-bacs@librecast.net\nSigned-off-by: Oscar Maes \u003coscmaes92@gmail.com\u003e\nReviewed-by: David Ahern \u003cdsahern@kernel.org\u003e\nLink: https://patch.msgid.link/20250827062322.4807-1-oscmaes92@gmail.com\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n\n" }, { "commit": "dac978e51cce0c1f00a14c4a82f81d387f79b2d4", "tree": "af2456b89b87a8761e093a58802800d82151e3d5", "parents": [ "bd2902e0bcac8e4daf9cf630517cda4448665667" ], "author": { "name": "Neil Mandir", "email": "neil.mandir@seco.com", "time": "Tue Aug 26 10:30:22 2025 -0400" }, "committer": { "name": "Paolo Abeni", "email": "pabeni@redhat.com", "time": "Thu Aug 28 10:12:21 2025 +0200" }, "message": "net: macb: Disable clocks once\n\nWhen the driver is removed the clocks are disabled twice: once in\nmacb_remove and a second time by runtime pm. Disable wakeup in remove so\nall the clocks are disabled and skip the second call to macb_clks_disable.\nAlways suspend the device as we always set it active in probe.\n\nFixes: d54f89af6cc4 (\"net: macb: Add pm runtime support\")\nSigned-off-by: Neil Mandir \u003cneil.mandir@seco.com\u003e\nCo-developed-by: Sean Anderson \u003csean.anderson@linux.dev\u003e\nSigned-off-by: Sean Anderson \u003csean.anderson@linux.dev\u003e\nLink: https://patch.msgid.link/20250826143022.935521-1-sean.anderson@linux.dev\nSigned-off-by: Paolo Abeni \u003cpabeni@redhat.com\u003e\n\n" }, { "commit": "a6358f8cf64850f3f27857b8ed8c1b08cfc4685c", "tree": "45e7f8d3fb6539bcb04f5b078eb1a3e5757b909a", "parents": [ "134ed1093907a79e5d98087513f13fd7652c4df9" ], "author": { "name": "Li Nan", "email": "linan122@huawei.com", "time": "Wed Aug 27 15:39:54 2025 +0800" }, "committer": { "name": "Ard Biesheuvel", "email": "ardb@kernel.org", "time": "Thu Aug 28 08:39:49 2025 +0200" }, "message": "efivarfs: Fix slab-out-of-bounds in efivarfs_d_compare\n\nObserved on kernel 6.6 (present on master as well):\n\n BUG: KASAN: slab-out-of-bounds in memcmp+0x98/0xd0\n Call trace:\n kasan_check_range+0xe8/0x190\n __asan_loadN+0x1c/0x28\n memcmp+0x98/0xd0\n efivarfs_d_compare+0x68/0xd8\n __d_lookup_rcu_op_compare+0x178/0x218\n __d_lookup_rcu+0x1f8/0x228\n d_alloc_parallel+0x150/0x648\n lookup_open.isra.0+0x5f0/0x8d0\n open_last_lookups+0x264/0x828\n path_openat+0x130/0x3f8\n do_filp_open+0x114/0x248\n do_sys_openat2+0x340/0x3c0\n __arm64_sys_openat+0x120/0x1a0\n\nIf dentry-\u003ed_name.len \u003c EFI_VARIABLE_GUID_LEN , \u0027guid\u0027 can become\nnegative, leadings to oob. The issue can be triggered by parallel\nlookups using invalid filename:\n\n T1\t\t\tT2\n lookup_open\n -\u003elookup\n simple_lookup\n d_add\n // invalid dentry is added to hash list\n\n\t\t\tlookup_open\n\t\t\t d_alloc_parallel\n\t\t\t __d_lookup_rcu\n\t\t\t __d_lookup_rcu_op_compare\n\t\t\t hlist_bl_for_each_entry_rcu\n\t\t\t // invalid dentry can be retrieved\n\t\t\t -\u003ed_compare\n\t\t\t efivarfs_d_compare\n\t\t\t // oob\n\nFix it by checking \u0027guid\u0027 before cmp.\n\nFixes: da27a24383b2 (\"efivarfs: guid part of filenames are case-insensitive\")\nSigned-off-by: Li Nan \u003clinan122@huawei.com\u003e\nSigned-off-by: Wu Guanghao \u003cwuguanghao3@huawei.com\u003e\nSigned-off-by: Ard Biesheuvel \u003cardb@kernel.org\u003e\n" }, { "commit": "82b8166171bdebbc74717e4a0cfb4b89cd0510aa", "tree": "35a50dc43cc599af1fea373b8271dcbf95c9ef37", "parents": [ "70c1595c181c48a022756116a6c46d5e8bad2c6f" ], "author": { "name": "Qianfeng Rong", "email": "rongqianfeng@vivo.com", "time": "Tue Aug 26 15:54:18 2025 +0800" }, "committer": { "name": "Damien Le Moal", "email": "dlemoal@kernel.org", "time": "Thu Aug 28 13:36:14 2025 +0900" }, "message": "ata: ahci_xgene: Use int type for \u0027rc\u0027 to store error codes\n\nUse int instead of u32 for the \u0027rc\u0027 variable in xgene_ahci_softreset()\nto store negative error codes returned by ahci_do_softreset().\n\nIn xgene_ahci_pmp_softreset(), remove the redundant \u0027rc\u0027 variable and\ndirectly return the result of the ahci_do_softreset() call instead.\n\nSigned-off-by: Qianfeng Rong \u003crongqianfeng@vivo.com\u003e\nSigned-off-by: Damien Le Moal \u003cdlemoal@kernel.org\u003e\n" }, { "commit": "07d9df80082b8d1f37e05658371b087cb6738770", "tree": "4044989053351c494427b7615d198cb8c7ce6834", "parents": [ "39f90c1967215375f7d87b81d14b0f3ed6b40c29", "2c369d91d0933aaff96b6b807b22363e6a38a625" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Aug 27 19:18:51 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Aug 27 19:18:51 2025 -0700" }, "message": "Merge tag \u0027perf-tools-fixes-for-v6.17-2025-08-27\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools\n\nPull perf-tools fixes from Namhyung Kim:\n \"A number of kernel header sync changes and two build-id fixes\"\n\n* tag \u0027perf-tools-fixes-for-v6.17-2025-08-27\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools:\n perf symbol: Add blocking argument to filename__read_build_id\n perf symbol-minimal: Fix ehdr reading in filename__read_build_id\n tools headers: Sync uapi/linux/vhost.h with the kernel source\n tools headers: Sync uapi/linux/prctl.h with the kernel source\n tools headers: Sync uapi/linux/fs.h with the kernel source\n tools headers: Sync uapi/linux/fcntl.h with the kernel source\n tools headers: Sync syscall tables with the kernel source\n tools headers: Sync powerpc headers with the kernel source\n tools headers: Sync arm64 headers with the kernel source\n tools headers: Sync x86 headers with the kernel source\n tools headers: Sync linux/cfi_types.h with the kernel source\n tools headers: Sync linux/bits.h with the kernel source\n tools headers: Sync KVM headers with the kernel source\n perf test: Fix a build error in x86 topdown test\n" }, { "commit": "bd2902e0bcac8e4daf9cf630517cda4448665667", "tree": "f87232479bf40b365a5c145ffab9a6840ae93f5d", "parents": [ "1cc8a5b534e5f9b5e129e54ee2e63c9f5da4f39a", "6ede14a2c6365e7e5d855643c7c8390b5268c467" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 18:57:13 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 18:57:13 2025 -0700" }, "message": "Merge branch \u0027locking-fixes-for-fbnic-driver\u0027\n\nAlexander Duyck says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nLocking fixes for fbnic driver\n\nAddress a few locking issues that were reported on the fbnic driver.\nSpecifically in one case we were seeing locking leaks due to us not\nreleasing the locks in certain exception paths. In another case we were\nusing phylink_resume outside of a section in which we held the RTNL mutex\nand as a result we were throwing an assert.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://patch.msgid.link/175616242563.1963577.7257712519613275567.stgit@ahduyck-xeon-server.home.arpa\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "6ede14a2c6365e7e5d855643c7c8390b5268c467", "tree": "f87232479bf40b365a5c145ffab9a6840ae93f5d", "parents": [ "2ddaa562b465921a5d1da3fc939993b92b953e20" ], "author": { "name": "Alexander Duyck", "email": "alexanderduyck@fb.com", "time": "Mon Aug 25 15:56:13 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 18:57:08 2025 -0700" }, "message": "fbnic: Move phylink resume out of service_task and into open/close\n\nThe fbnic driver was presenting with the following locking assert coming\nout of a PM resume:\n[ 42.208116][ T164] RTNL: assertion failed at drivers/net/phy/phylink.c (2611)\n[ 42.208492][ T164] WARNING: CPU: 1 PID: 164 at drivers/net/phy/phylink.c:2611 phylink_resume+0x190/0x1e0\n[ 42.208872][ T164] Modules linked in:\n[ 42.209140][ T164] CPU: 1 UID: 0 PID: 164 Comm: bash Not tainted 6.17.0-rc2-virtme #134 PREEMPT(full)\n[ 42.209496][ T164] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.17.0-5.fc42 04/01/2014\n[ 42.209861][ T164] RIP: 0010:phylink_resume+0x190/0x1e0\n[ 42.210057][ T164] Code: 83 e5 01 0f 85 b0 fe ff ff c6 05 1c cd 3e 02 01 90 ba 33 0a 00 00 48 c7 c6 20 3a 1d a5 48 c7 c7 e0 3e 1d a5 e8 21 b8 90 fe 90 \u003c0f\u003e 0b 90 90 e9 86 fe ff ff e8 42 ea 1f ff e9 e2 fe ff ff 48 89 ef\n[ 42.210708][ T164] RSP: 0018:ffffc90000affbd8 EFLAGS: 00010296\n[ 42.210983][ T164] RAX: 0000000000000000 RBX: ffff8880078d8400 RCX: 0000000000000000\n[ 42.211235][ T164] RDX: 0000000000000000 RSI: 1ffffffff4f10938 RDI: 0000000000000001\n[ 42.211466][ T164] RBP: 0000000000000000 R08: ffffffffa2ae79ea R09: fffffbfff4b3eb84\n[ 42.211707][ T164] R10: 0000000000000003 R11: 0000000000000000 R12: ffff888007ad8000\n[ 42.211997][ T164] R13: 0000000000000002 R14: ffff888006a18800 R15: ffffffffa34c59e0\n[ 42.212234][ T164] FS: 00007f0dc8e39740(0000) GS:ffff88808f51f000(0000) knlGS:0000000000000000\n[ 42.212505][ T164] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033\n[ 42.212704][ T164] CR2: 00007f0dc8e9fe10 CR3: 000000000b56d003 CR4: 0000000000772ef0\n[ 42.213227][ T164] PKRU: 55555554\n[ 42.213366][ T164] Call Trace:\n[ 42.213483][ T164] \u003cTASK\u003e\n[ 42.213565][ T164] __fbnic_pm_attach.isra.0+0x8e/0xa0\n[ 42.213725][ T164] pci_reset_function+0x116/0x1d0\n[ 42.213895][ T164] reset_store+0xa0/0x100\n[ 42.214025][ T164] ? pci_dev_reset_attr_is_visible+0x50/0x50\n[ 42.214221][ T164] ? sysfs_file_kobj+0xc1/0x1e0\n[ 42.214374][ T164] ? sysfs_kf_write+0x65/0x160\n[ 42.214526][ T164] kernfs_fop_write_iter+0x2f8/0x4c0\n[ 42.214677][ T164] ? kernfs_vma_page_mkwrite+0x1f0/0x1f0\n[ 42.214836][ T164] new_sync_write+0x308/0x6f0\n[ 42.214987][ T164] ? __lock_acquire+0x34c/0x740\n[ 42.215135][ T164] ? new_sync_read+0x6f0/0x6f0\n[ 42.215288][ T164] ? lock_acquire.part.0+0xbc/0x260\n[ 42.215440][ T164] ? ksys_write+0xff/0x200\n[ 42.215590][ T164] ? perf_trace_sched_switch+0x6d0/0x6d0\n[ 42.215742][ T164] vfs_write+0x65e/0xbb0\n[ 42.215876][ T164] ksys_write+0xff/0x200\n[ 42.215994][ T164] ? __ia32_sys_read+0xc0/0xc0\n[ 42.216141][ T164] ? do_user_addr_fault+0x269/0x9f0\n[ 42.216292][ T164] ? rcu_is_watching+0x15/0xd0\n[ 42.216442][ T164] do_syscall_64+0xbb/0x360\n[ 42.216591][ T164] entry_SYSCALL_64_after_hwframe+0x4b/0x53\n[ 42.216784][ T164] RIP: 0033:0x7f0dc8ea9986\n\nA bit of digging showed that we were invoking the phylink_resume as a part\nof the fbnic_up path when we were enabling the service task while not\nholding the RTNL lock. We should be enabling this sooner as a part of the\nndo_open path and then just letting the service task come online later.\nThis will help to enforce the correct locking and brings the phylink\ninterface online at the same time as the network interface, instead of at a\nlater time.\n\nI tested this on QEMU to verify this was working by putting the system to\nsleep using \"echo mem \u003e /sys/power/state\" to put the system to sleep in the\nguest and then using the command \"system_wakeup\" in the QEMU monitor.\n\nFixes: 69684376eed5 (\"eth: fbnic: Add link detection\")\nSigned-off-by: Alexander Duyck \u003calexanderduyck@fb.com\u003e\nReviewed-by: Przemek Kitszel \u003cprzemyslaw.kitszel@intel.com\u003e\nLink: https://patch.msgid.link/175616257316.1963577.12238158800417771119.stgit@ahduyck-xeon-server.home.arpa\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "2ddaa562b465921a5d1da3fc939993b92b953e20", "tree": "2c9576672999699524864a67a9e84c304bcc8422", "parents": [ "1cc8a5b534e5f9b5e129e54ee2e63c9f5da4f39a" ], "author": { "name": "Alexander Duyck", "email": "alexanderduyck@fb.com", "time": "Mon Aug 25 15:56:06 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 18:57:08 2025 -0700" }, "message": "fbnic: Fixup rtnl_lock and devl_lock handling related to mailbox code\n\nThe exception handling path for the __fbnic_pm_resume function had a bug in\nthat it was taking the devlink lock and then exiting to exception handling\ninstead of waiting until after it released the lock to do so. In order to\nhandle that I am swapping the placement of the unlock and the exception\nhandling jump to label so that we don\u0027t trigger a deadlock by holding the\nlock longer than we need to.\n\nIn addition this change applies the same ordering to the rtnl_lock/unlock\ncalls in the same function as it should make the code easier to follow if\nit adheres to a consistent pattern.\n\nFixes: 82534f446daa (\"eth: fbnic: Add devlink dev flash support\")\nSigned-off-by: Alexander Duyck \u003calexanderduyck@fb.com\u003e\nReviewed-by: Przemek Kitszel \u003cprzemyslaw.kitszel@intel.com\u003e\nLink: https://patch.msgid.link/175616256667.1963577.5543500806256052549.stgit@ahduyck-xeon-server.home.arpa\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "1cc8a5b534e5f9b5e129e54ee2e63c9f5da4f39a", "tree": "523a4c06577608a53108198071e1f0f5d325f7f4", "parents": [ "9b8c88f875c04d4cb9111bd5dd9291c7e9691bf5" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Wed Aug 27 17:21:49 2025 +0000" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 17:27:52 2025 -0700" }, "message": "net: rose: fix a typo in rose_clear_routes()\n\nsyzbot crashed in rose_clear_routes(), after a recent patch typo.\n\nKASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017]\nCPU: 0 UID: 0 PID: 10591 Comm: syz.3.1856 Not tainted syzkaller #0 PREEMPT(full)\nHardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 07/12/2025\n RIP: 0010:rose_clear_routes net/rose/rose_route.c:565 [inline]\n RIP: 0010:rose_rt_ioctl+0x162/0x1250 net/rose/rose_route.c:760\n \u003cTASK\u003e\n rose_ioctl+0x3ce/0x8b0 net/rose/af_rose.c:1381\n sock_do_ioctl+0xd9/0x300 net/socket.c:1238\n sock_ioctl+0x576/0x790 net/socket.c:1359\n vfs_ioctl fs/ioctl.c:51 [inline]\n __do_sys_ioctl fs/ioctl.c:598 [inline]\n __se_sys_ioctl+0xfc/0x170 fs/ioctl.c:584\n do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]\n do_syscall_64+0xfa/0x3b0 arch/x86/entry/syscall_64.c:94\n entry_SYSCALL_64_after_hwframe+0x77/0x7f\n\nFixes: da9c9c877597 (\"net: rose: include node references in rose_neigh refcount\")\nReported-by: syzbot+2eb8d1719f7cfcfa6840@syzkaller.appspotmail.com\nCloses: https://lore.kernel.org/netdev/68af3e29.a70a0220.3cafd4.002e.GAE@google.com/T/#u\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: Takamitsu Iwai \u003ctakamitz@amazon.co.jp\u003e\nReviewed-by: Kuniyuki Iwashima \u003ckuniyu@google.com\u003e\nLink: https://patch.msgid.link/20250827172149.5359-1-edumazet@google.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "9b8c88f875c04d4cb9111bd5dd9291c7e9691bf5", "tree": "94e9cc26a55f0fd899424d591835818c1de274b8", "parents": [ "2e8750469242cad8f01f320131fd5a6f540dbb99" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Tue Aug 26 13:44:35 2025 +0000" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 17:16:13 2025 -0700" }, "message": "l2tp: do not use sock_hold() in pppol2tp_session_get_sock()\n\npppol2tp_session_get_sock() is using RCU, it must be ready\nfor sk_refcnt being zero.\n\nCommit ee40fb2e1eb5 (\"l2tp: protect sock pointer of\nstruct pppol2tp_session with RCU\") was correct because it\nhad a call_rcu(..., pppol2tp_put_sk) which was later removed in blamed commit.\n\npppol2tp_recv() can use pppol2tp_session_get_sock() as well.\n\nFixes: c5cbaef992d6 (\"l2tp: refactor ppp socket/session relationship\")\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: James Chapman \u003cjchapman@katalix.com\u003e\nReviewed-by: Guillaume Nault \u003cgnault@redhat.com\u003e\nLink: https://patch.msgid.link/20250826134435.1683435-1-edumazet@google.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "2e8750469242cad8f01f320131fd5a6f540dbb99", "tree": "ad7e94a84ec3c5ce3314741af6c0d06e25e5c459", "parents": [ "bcd6f8954dc4a3aa32edda5602e43a0174dc8f0f" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Tue Aug 26 14:13:14 2025 +0000" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 17:15:21 2025 -0700" }, "message": "sctp: initialize more fields in sctp_v6_from_sk()\n\nsyzbot found that sin6_scope_id was not properly initialized,\nleading to undefined behavior.\n\nClear sin6_scope_id and sin6_flowinfo.\n\nBUG: KMSAN: uninit-value in __sctp_v6_cmp_addr+0x887/0x8c0 net/sctp/ipv6.c:649\n __sctp_v6_cmp_addr+0x887/0x8c0 net/sctp/ipv6.c:649\n sctp_inet6_cmp_addr+0x4f2/0x510 net/sctp/ipv6.c:983\n sctp_bind_addr_conflict+0x22a/0x3b0 net/sctp/bind_addr.c:390\n sctp_get_port_local+0x21eb/0x2440 net/sctp/socket.c:8452\n sctp_get_port net/sctp/socket.c:8523 [inline]\n sctp_listen_start net/sctp/socket.c:8567 [inline]\n sctp_inet_listen+0x710/0xfd0 net/sctp/socket.c:8636\n __sys_listen_socket net/socket.c:1912 [inline]\n __sys_listen net/socket.c:1927 [inline]\n __do_sys_listen net/socket.c:1932 [inline]\n __se_sys_listen net/socket.c:1930 [inline]\n __x64_sys_listen+0x343/0x4c0 net/socket.c:1930\n x64_sys_call+0x271d/0x3e20 arch/x86/include/generated/asm/syscalls_64.h:51\n do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]\n do_syscall_64+0xd9/0x210 arch/x86/entry/syscall_64.c:94\n entry_SYSCALL_64_after_hwframe+0x77/0x7f\n\nLocal variable addr.i.i created at:\n sctp_get_port net/sctp/socket.c:8515 [inline]\n sctp_listen_start net/sctp/socket.c:8567 [inline]\n sctp_inet_listen+0x650/0xfd0 net/sctp/socket.c:8636\n __sys_listen_socket net/socket.c:1912 [inline]\n __sys_listen net/socket.c:1927 [inline]\n __do_sys_listen net/socket.c:1932 [inline]\n __se_sys_listen net/socket.c:1930 [inline]\n __x64_sys_listen+0x343/0x4c0 net/socket.c:1930\n\nFixes: 1da177e4c3f4 (\"Linux-2.6.12-rc2\")\nReported-by: syzbot+e69f06a0f30116c68056@syzkaller.appspotmail.com\nCloses: https://lore.kernel.org/netdev/68adc0a2.050a0220.37038e.00c4.GAE@google.com/T/#u\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: Marcelo Ricardo Leitner \u003cmarcelo.leitner@gmail.com\u003e\nAcked-by: Xin Long \u003clucien.xin@gmail.com\u003e\nLink: https://patch.msgid.link/20250826141314.1802610-1-edumazet@google.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "bcd6f8954dc4a3aa32edda5602e43a0174dc8f0f", "tree": "e19a368a14c12385f452d1eb4c17d0a1edbc5712", "parents": [ "ceb9515524046252c522b16f38881e8837ec0d91" ], "author": { "name": "Subash Abhinov Kasiviswanathan", "email": "subash.a.kasiviswanathan@oss.qualcomm.com", "time": "Tue Aug 26 14:50:46 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 17:08:56 2025 -0700" }, "message": "MAINTAINERS: rmnet: Update email addresses\n\nSwitch to oss.qualcomm.com ids.\n\nSigned-off-by: Sean Tranchetti \u003csean.tranchetti@oss.qualcomm.com\u003e\nSigned-off-by: Subash Abhinov Kasiviswanathan \u003csubash.a.kasiviswanathan@oss.qualcomm.com\u003e\nLink: https://patch.msgid.link/20250826215046.865530-1-subash.a.kasiviswanathan@oss.qualcomm.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "ab529e6ca1f67bcf31f3ea80c72bffde2e9e053e", "tree": "9ec513c082fbbee48a0e6e242a12594ce2cf0692", "parents": [ "0e08fa789d39aa01923e3ba144bd808291895c3c" ], "author": { "name": "Shuhao Fu", "email": "sfual@cse.ust.hk", "time": "Thu Aug 28 02:24:19 2025 +0800" }, "committer": { "name": "Steve French", "email": "stfrench@microsoft.com", "time": "Wed Aug 27 14:59:06 2025 -0500" }, "message": "fs/smb: Fix inconsistent refcnt update\n\nA possible inconsistent update of refcount was identified in `smb2_compound_op`.\nSuch inconsistent update could lead to possible resource leaks.\n\nWhy it is a possible bug:\n1. In the comment section of the function, it clearly states that the\nreference to `cfile` should be dropped after calling this function.\n2. Every control flow path would check and drop the reference to\n`cfile`, except the patched one.\n3. Existing callers would not handle refcount update of `cfile` if\n-ENOMEM is returned.\n\nTo fix the bug, an extra goto label \"out\" is added, to make sure that the\ncleanup logic would always be respected. As the problem is caused by the\nallocation failure of `vars`, the cleanup logic between label \"finished\"\nand \"out\" can be safely ignored. According to the definition of function\n`is_replayable_error`, the error code of \"-ENOMEM\" is not recoverable.\nTherefore, the replay logic also gets ignored.\n\nSigned-off-by: Shuhao Fu \u003csfual@cse.ust.hk\u003e\nAcked-by: Paulo Alcantara (Red Hat) \u003cpc@manguebit.org\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Steve French \u003cstfrench@microsoft.com\u003e\n" }, { "commit": "c767d74a9cdd1042046d02319d16b85d9aa8a8aa", "tree": "1939c216e0091cb6eb781a7bcc59d281b9793e5f", "parents": [ "ee38ea0ae4ed13fe33e033dc98d11e76bc7167cd" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Fri Aug 22 12:12:37 2025 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Aug 27 14:01:52 2025 -0400" }, "message": "drm/amdgpu/userq: fix error handling of invalid doorbell\n\nIf the doorbell is invalid, be sure to set the r to an error\nstate so the function returns an error.\n\nReviewed-by: David (Ming Qiang) Wu \u003cDavid.Wu3@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n(cherry picked from commit 7e2a5b0a9a165a7c51274aa01b18be29491b4345)\nCc: stable@vger.kernel.org\n" }, { "commit": "ee38ea0ae4ed13fe33e033dc98d11e76bc7167cd", "tree": "dbfd8425fb60a52c531e15d5916f44df08a77b92", "parents": [ "5dff50802b285da8284a7bf17ae2fdc6f1357023" ], "author": { "name": "Jesse.Zhang", "email": "Jesse.Zhang@amd.com", "time": "Tue Aug 26 17:30:58 2025 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Aug 27 14:01:32 2025 -0400" }, "message": "drm/amdgpu: update firmware version checks for user queue support\n\nThe minimum firmware versions required for user queue functionality\nhave been increased to address an issue where the queue privilege\nstate was lost during queue connect operations.\n\nThe problem occurred because the privilege state was being restored\nto its initial value at the beginning of the function, overwriting\nthe state that was properly set during the queue connect case.\n\nThis commit updates the minimum version requirements:\n- ME firmware from 2390 to 2420\n- PFP firmware from 2530 to 2580\n- MEC firmware from 2600 to 2650\n- MES firmware remains at 120\n\nThese updated firmware versions contain the necessary fixes to\nproperly maintain queue privilege state throughout connect operations.\n\nFixes: 61ca97e9590c (\"drm/amdgpu: Add fw minimum version check for usermode queue\")\nAcked-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nSigned-off-by: Jesse Zhang \u003cJesse.Zhang@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n(cherry picked from commit 5f976c9939f0d5916d2b8ef3156a6d1799781df1)\nCc: stable@vger.kernel.org\n" }, { "commit": "5dff50802b285da8284a7bf17ae2fdc6f1357023", "tree": "8cd5f48146edbd1729a987cd001d1f8cc1f3ea94", "parents": [ "ac4ed2da4c1305a1a002415058aa7deaf49ffe3e" ], "author": { "name": "Yang Wang", "email": "kevinyang.wang@amd.com", "time": "Mon Aug 25 12:54:01 2025 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Aug 27 14:01:08 2025 -0400" }, "message": "drm/amd/amdgpu: disable hwmon power1_cap* for gfx 11.0.3 on vf mode\n\nthe PPSMC_MSG_GetPptLimit msg is not valid for gfx 11.0.3 on vf mode,\nso skiped to create power1_cap* hwmon sysfs node.\n\nSigned-off-by: Yang Wang \u003ckevinyang.wang@amd.com\u003e\nReviewed-by: Asad Kamal \u003casad.kamal@amd.com\u003e\nAcked-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n(cherry picked from commit e82a8d441038d8cb10b63047a9e705c42479d156)\nCc: stable@vger.kernel.org\n" }, { "commit": "ac4ed2da4c1305a1a002415058aa7deaf49ffe3e", "tree": "124dcc04035ba43afe887db13bbe0a1503b14682", "parents": [ "29f155c5e82fe35ff85b1f13612cb8c2dbe1dca3" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Mon Aug 25 13:40:22 2025 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Aug 27 14:00:38 2025 -0400" }, "message": "Revert \"drm/amdgpu: fix incorrect vm flags to map bo\"\n\nThis reverts commit b08425fa77ad2f305fe57a33dceb456be03b653f.\n\nRevert this to align with 6.17 because the fixes tag\nwas wrong on this commit.\n\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n(cherry picked from commit be33e8a239aac204d7e9e673c4220ef244eb1ba3)\n" }, { "commit": "29f155c5e82fe35ff85b1f13612cb8c2dbe1dca3", "tree": "0fdb34a2e76a31e368c8a26f4e5731a2276c0034", "parents": [ "27f5e0c1321ee280189cea16044de2e157dc4bb9" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Jun 24 11:38:14 2025 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Aug 27 13:59:53 2025 -0400" }, "message": "drm/amdgpu/gfx12: set MQD as appriopriate for queue types\n\nSet the MQD as appropriate for the kernel vs user queues.\n\nAcked-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Lijo Lazar \u003clijo.lazar@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n(cherry picked from commit 7b9110f2897957efd9715b52fc01986509729db3)\nCc: stable@vger.kernel.org\n" }, { "commit": "27f5e0c1321ee280189cea16044de2e157dc4bb9", "tree": "43d768a4dafc12f5aa16389d0169d5b20ad4d431", "parents": [ "1b237f190eb3d36f52dffe07a40b5eb210280e00" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Tue Jun 24 11:37:16 2025 -0400" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Aug 27 13:59:25 2025 -0400" }, "message": "drm/amdgpu/gfx11: set MQD as appriopriate for queue types\n\nSet the MQD as appropriate for the kernel vs user queues.\n\nAcked-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Lijo Lazar \u003clijo.lazar@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n(cherry picked from commit 063d6683208722b1875f888a45084e3d112701ac)\nCc: stable@vger.kernel.org\n" }, { "commit": "39f90c1967215375f7d87b81d14b0f3ed6b40c29", "tree": "6f82b5978f946f2117d2e7dff4a0a25cc0633e1d", "parents": [ "518b21ba139cefa2ee7f9fcf516fdc6743e8db68", "45d8ef6322b8a828d3b1e2cfb8893e2ff882cb23" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Aug 27 10:19:35 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Aug 27 10:19:35 2025 -0700" }, "message": "Merge tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost\n\nPull virtio/vhost fixes from Michael Tsirkin:\n \"More small fixes. Most notably this fixes a messed up ioctl number,\n and a regression in shmem affecting drm users\"\n\n* tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:\n virtio_net: adjust the execution order of function `virtnet_close` during freeze\n virtio_input: Improve freeze handling\n vhost: Fix ioctl # for VHOST_[GS]ET_FORK_FROM_OWNER\n Revert \"virtio: reject shm region if length is zero\"\n vhost/net: Protect ubufs with rcu read lock in vhost_net_ubuf_put()\n virtio_pci: Fix misleading comment for queue vector\n" }, { "commit": "518b21ba139cefa2ee7f9fcf516fdc6743e8db68", "tree": "db20e67af19d1d43b55769ca6778bdcf1d09d6e4", "parents": [ "fab1beda7597fac1cecc01707d55eadb6bbe773c", "6f6fbd9a0c5a75eee0618c1499cf73cc770b3f52" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Aug 27 10:10:50 2025 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Aug 27 10:10:50 2025 -0700" }, "message": "Merge tag \u0027media/v6.17-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media\n\nPull media fixes from Mauro Carvalho Chehab:\n\n - drop the redundant pm_runtime_mark_last_busy() in rkvdec\n\n - fix probing error handling in rkvdec\n\n - fix an issue affecting lt6911uxe/lt6911uxc related to CSI-2 GPIO pins\n in int3472\n\n* tag \u0027media/v6.17-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:\n media: Remove redundant pm_runtime_mark_last_busy() calls\n platform/x86: int3472: add hpd pin support\n media: rkvdec: Remove redundant pm_runtime_mark_last_busy() calls\n media: rkvdec: Fix an error handling path in rkvdec_probe()\n media: rkvdec: Fix a NULL vs IS_ERR() bug in probe()\n" }, { "commit": "ceb9515524046252c522b16f38881e8837ec0d91", "tree": "4fdf9fce70085fc267ea45c4982a6f101685766d", "parents": [ "9448ccd853368582efa9db05db344f8bb9dffe0f", "da9c9c877597170b929a6121a68dcd3dd9a80f45" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 07:43:20 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 07:43:20 2025 -0700" }, "message": "Merge branch \u0027introduce-refcount_t-for-reference-counting-of-rose_neigh\u0027\n\nTakamitsu Iwai says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nIntroduce refcount_t for reference counting of rose_neigh\n\nThe current implementation of rose_neigh uses \u0027use\u0027 and \u0027count\u0027 field of\ntype unsigned short as a reference count. This approach lacks atomicity,\nleading to potential race conditions. As a result, syzbot has reported\nslab-use-after-free errors due to unintended removals.\n\nThis series introduces refcount_t for reference counting to ensure\natomicity and prevent race conditions. The patches are structured as\nfollows:\n\n1. Refactor rose_remove_neigh() to separate removal and freeing operations\n2. Convert \u0027use\u0027 field to refcount_t for appropriate reference counting\n3. Include references from rose_node to \u0027use\u0027 field\n\nThese changes should resolve the reported slab-use-after-free issues and\nimprove the overall stability of the ROSE network layer.\n\nv1: https://lore.kernel.org/20250820174707.83372-1-takamitz@amazon.co.jp\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://patch.msgid.link/20250823085857.47674-1-takamitz@amazon.co.jp\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "da9c9c877597170b929a6121a68dcd3dd9a80f45", "tree": "4fdf9fce70085fc267ea45c4982a6f101685766d", "parents": [ "d860d1faa6b2ce3becfdb8b0c2b048ad31800061" ], "author": { "name": "Takamitsu Iwai", "email": "takamitz@amazon.co.jp", "time": "Sat Aug 23 17:58:57 2025 +0900" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 07:43:09 2025 -0700" }, "message": "net: rose: include node references in rose_neigh refcount\n\nCurrent implementation maintains two separate reference counting\nmechanisms: the \u0027count\u0027 field in struct rose_neigh tracks references from\nrose_node structures, while the \u0027use\u0027 field (now refcount_t) tracks\nreferences from rose_sock.\n\nThis patch merges these two reference counting systems using \u0027use\u0027 field\nfor proper reference management. Specifically, this patch adds incrementing\nand decrementing of rose_neigh-\u003euse when rose_neigh-\u003ecount is incremented\nor decremented.\n\nThis patch also modifies rose_rt_free(), rose_rt_device_down() and\nrose_clear_route() to properly release references to rose_neigh objects\nbefore freeing a rose_node through rose_remove_node().\n\nThese changes ensure rose_neigh structures are properly freed only when\nall references, including those from rose_node structures, are released.\nAs a result, this resolves a slab-use-after-free issue reported by Syzbot.\n\nFixes: 1da177e4c3f4 (\"Linux-2.6.12-rc2\")\nReported-by: syzbot+942297eecf7d2d61d1f1@syzkaller.appspotmail.com\nCloses: https://syzkaller.appspot.com/bug?extid\u003d942297eecf7d2d61d1f1\nSigned-off-by: Takamitsu Iwai \u003ctakamitz@amazon.co.jp\u003e\nReviewed-by: Kuniyuki Iwashima \u003ckuniyu@google.com\u003e\nLink: https://patch.msgid.link/20250823085857.47674-4-takamitz@amazon.co.jp\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "d860d1faa6b2ce3becfdb8b0c2b048ad31800061", "tree": "a27770249cfd5066696f5bc917acf37170a632d1", "parents": [ "dcb34659028f856c423a29ef9b4e2571d203444d" ], "author": { "name": "Takamitsu Iwai", "email": "takamitz@amazon.co.jp", "time": "Sat Aug 23 17:58:56 2025 +0900" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 07:43:08 2025 -0700" }, "message": "net: rose: convert \u0027use\u0027 field to refcount_t\n\nThe \u0027use\u0027 field in struct rose_neigh is used as a reference counter but\nlacks atomicity. This can lead to race conditions where a rose_neigh\nstructure is freed while still being referenced by other code paths.\n\nFor example, when rose_neigh-\u003euse becomes zero during an ioctl operation\nvia rose_rt_ioctl(), the structure may be removed while its timer is\nstill active, potentially causing use-after-free issues.\n\nThis patch changes the type of \u0027use\u0027 from unsigned short to refcount_t and\nupdates all code paths to use rose_neigh_hold() and rose_neigh_put() which\noperate reference counts atomically.\n\nFixes: 1da177e4c3f4 (\"Linux-2.6.12-rc2\")\nSigned-off-by: Takamitsu Iwai \u003ctakamitz@amazon.co.jp\u003e\nReviewed-by: Kuniyuki Iwashima \u003ckuniyu@google.com\u003e\nLink: https://patch.msgid.link/20250823085857.47674-3-takamitz@amazon.co.jp\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "dcb34659028f856c423a29ef9b4e2571d203444d", "tree": "d4c9e40f6587fd8438ecbc09b851522ea6a08484", "parents": [ "9448ccd853368582efa9db05db344f8bb9dffe0f" ], "author": { "name": "Takamitsu Iwai", "email": "takamitz@amazon.co.jp", "time": "Sat Aug 23 17:58:55 2025 +0900" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Wed Aug 27 07:43:08 2025 -0700" }, "message": "net: rose: split remove and free operations in rose_remove_neigh()\n\nThe current rose_remove_neigh() performs two distinct operations:\n1. Removes rose_neigh from rose_neigh_list\n2. Frees the rose_neigh structure\n\nSplit these operations into separate functions to improve maintainability\nand prepare for upcoming refcount_t conversion. The timer cleanup remains\nin rose_remove_neigh() because free operations can be called from timer\nitself.\n\nThis patch introduce rose_neigh_put() to handle the freeing of rose_neigh\nstructures and modify rose_remove_neigh() to handle removal only.\n\nSigned-off-by: Takamitsu Iwai \u003ctakamitz@amazon.co.jp\u003e\nReviewed-by: Kuniyuki Iwashima \u003ckuniyu@google.com\u003e\nLink: https://patch.msgid.link/20250823085857.47674-2-takamitz@amazon.co.jp\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "c64eff368ac676e8540344d27a3de47e0ad90d21", "tree": "089c55d83a80667ce71b96721acfe9a0d5eedd44", "parents": [ "e4e6aaea46b7be818eba0510ba68d30df8689ea3" ], "author": { "name": "Qingyue Zhang", "email": "chunzhennn@qq.com", "time": "Wed Aug 27 19:43:39 2025 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Wed Aug 27 08:39:44 2025 -0600" }, "message": "io_uring/kbuf: fix signedness in this_len calculation\n\nWhen importing and using buffers, buf-\u003elen is considered unsigned.\nHowever, buf-\u003elen is converted to signed int when committing. This can\nlead to unexpected behavior if the buffer is large enough to be\ninterpreted as a negative value. Make min_t calculation unsigned.\n\nFixes: ae98dbf43d75 (\"io_uring/kbuf: add support for incremental buffer consumption\")\nCo-developed-by: Suoxing Zhang \u003caftern00n@qq.com\u003e\nSigned-off-by: Suoxing Zhang \u003caftern00n@qq.com\u003e\nSigned-off-by: Qingyue Zhang \u003cchunzhennn@qq.com\u003e\nLink: https://lore.kernel.org/r/tencent_4DBB3674C0419BEC2C0C525949DA410CA307@qq.com\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "22b2ca023fc41e95afb3c58c16c2418b1ca85d75", "tree": "bf3b8585d8c309ff40fedd3dc299d3c34a207e0c", "parents": [ "1b237f190eb3d36f52dffe07a40b5eb210280e00", "dce1b33ed7430c7189b8cc1567498f9e6bf12731" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Wed Aug 27 04:18:01 2025 -0400" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Wed Aug 27 04:18:01 2025 -0400" }, "message": "Merge tag \u0027kvm-x86-fixes-6.17-rc7\u0027 of https://github.com/kvm-x86/linux into HEAD\n\nKVM x86 fixes and a selftest fix for 6.17-rcN\n\n - Use array_index_nospec() to sanitize the target vCPU ID when handling PV\n IPIs and yields as the ID is guest-controlled.\n\n - Drop a superfluous cpumask_empty() check when reclaiming SEV memory, as\n the common case, by far, is that at least one CPU will have entered the\n VM, and wbnoinvd_on_cpus_mask() will naturally handle the rare case where\n the set of have_run_cpus is empty.\n\n - Rename the is_signed_type() macro in kselftest_harness.h to is_signed_var()\n to fix a collision with linux/overflow.h. The collision generates compiler\n warnings due to the two macros having different implementations.\n" }, { "commit": "9448ccd853368582efa9db05db344f8bb9dffe0f", "tree": "b77bc7fb4d8a7c2291ab05434626bf6d855e73d5", "parents": [ "4b4a8ec21ab8581f57d747b8e7eac65d3c295cd4" ], "author": { "name": "Dipayaan Roy", "email": "dipayanroy@linux.microsoft.com", "time": "Mon Aug 25 04:56:27 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:15:19 2025 -0700" }, "message": "net: hv_netvsc: fix loss of early receive events from host during channel open.\n\nThe hv_netvsc driver currently enables NAPI after opening the primary and\nsubchannels. This ordering creates a race: if the Hyper-V host places data\nin the host -\u003e guest ring buffer and signals the channel before\nnapi_enable() has been called, the channel callback will run but\nnapi_schedule_prep() will return false. As a result, the NAPI poller never\ngets scheduled, the data in the ring buffer is not consumed, and the\nreceive queue may remain permanently stuck until another interrupt happens\nto arrive.\n\nFix this by enabling NAPI and registering it with the RX/TX queues before\nvmbus channel is opened. This guarantees that any early host signal after\nopen will correctly trigger NAPI scheduling and the ring buffer will be\ndrained.\n\nFixes: 76bb5db5c749d (\"netvsc: fix use after free on module removal\")\nSigned-off-by: Dipayaan Roy \u003cdipayanroy@linux.microsoft.com\u003e\nLink: https://patch.msgid.link/20250825115627.GA32189@linuxonhyperv3.guj3yctzbm1etfxqx2vob5hsef.xx.internal.cloudapp.net\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "4b4a8ec21ab8581f57d747b8e7eac65d3c295cd4", "tree": "5e47771eb2d2004bdea3cb6aa453c4e071f0b621", "parents": [ "24be0900478dd87b945c7809e343fa07888a24b4", "b1eded580ab28119de0b0f21efe37ee2b4419144" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:12:45 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:12:46 2025 -0700" }, "message": "Merge branch \u0027net-stmmac-xgmac-minor-fixes\u0027\n\nRohan G Thomas says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nnet: stmmac: xgmac: Minor fixes\n\nThis patch series includes following minor fixes for stmmac\ndwxgmac driver:\n\n 1. Disable Rx FIFO overflow interrupt for dwxgmac\n 2. Correct supported speed modes for dwxgmac\n 3. Check for coe-unsupported flag before setting CIC bit of\n Tx Desc3 in the AF_XDP flow\n\nv2: https://lore.kernel.org/20250816-xgmac-minor-fixes-v2-0-699552cf8a7f@altera.com\nv1: https://lore.kernel.org/20250714-xgmac-minor-fixes-v1-0-c34092a88a72@altera.com\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://patch.msgid.link/20250825-xgmac-minor-fixes-v3-0-c225fe4444c0@altera.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "b1eded580ab28119de0b0f21efe37ee2b4419144", "tree": "5e47771eb2d2004bdea3cb6aa453c4e071f0b621", "parents": [ "42ef11b2bff5b6a2910c28d2ea47cc00e0fbcaec" ], "author": { "name": "Rohan G Thomas", "email": "rohan.g.thomas@altera.com", "time": "Mon Aug 25 12:36:54 2025 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:12:42 2025 -0700" }, "message": "net: stmmac: Set CIC bit only for TX queues with COE\n\nCurrently, in the AF_XDP transmit paths, the CIC bit of\nTX Desc3 is set for all packets. Setting this bit for\npackets transmitting through queues that don\u0027t support\nchecksum offloading causes the TX DMA to get stuck after\ntransmitting some packets. This patch ensures the CIC bit\nof TX Desc3 is set only if the TX queue supports checksum\noffloading.\n\nFixes: 132c32ee5bc0 (\"net: stmmac: Add TX via XDP zero-copy socket\")\nSigned-off-by: Rohan G Thomas \u003crohan.g.thomas@altera.com\u003e\nReviewed-by: Matthew Gerlach \u003cmatthew.gerlach@altera.com\u003e\nLink: https://patch.msgid.link/20250825-xgmac-minor-fixes-v3-3-c225fe4444c0@altera.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "42ef11b2bff5b6a2910c28d2ea47cc00e0fbcaec", "tree": "dfdce458ed17092f680f1b916ea81bcf0d97543c", "parents": [ "4f23382841e67174211271a454811dd17c0ef3c5" ], "author": { "name": "Rohan G Thomas", "email": "rohan.g.thomas@altera.com", "time": "Mon Aug 25 12:36:53 2025 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:12:42 2025 -0700" }, "message": "net: stmmac: xgmac: Correct supported speed modes\n\nCorrect supported speed modes as per the XGMAC databook.\nCommit 9cb54af214a7 (\"net: stmmac: Fix IP-cores specific\nMAC capabilities\") removes support for 10M, 100M and\n1000HD. 1000HD is not supported by XGMAC IP, but it does\nsupport 10M and 100M FD mode for XGMAC version \u003e\u003d 2_20,\nand it also supports 10M and 100M HD mode if the HDSEL bit\nis set in the MAC_HW_FEATURE0 reg. This commit enables support\nfor 10M and 100M speed modes for XGMAC IP based on XGMAC\nversion and MAC capabilities.\n\nFixes: 9cb54af214a7 (\"net: stmmac: Fix IP-cores specific MAC capabilities\")\nSigned-off-by: Rohan G Thomas \u003crohan.g.thomas@altera.com\u003e\nReviewed-by: Matthew Gerlach \u003cmatthew.gerlach@altera.com\u003e\nLink: https://patch.msgid.link/20250825-xgmac-minor-fixes-v3-2-c225fe4444c0@altera.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "4f23382841e67174211271a454811dd17c0ef3c5", "tree": "f0f0d3bcafdf2d5fe502ba1fa932dd8179d0b985", "parents": [ "24be0900478dd87b945c7809e343fa07888a24b4" ], "author": { "name": "Rohan G Thomas", "email": "rohan.g.thomas@altera.com", "time": "Mon Aug 25 12:36:52 2025 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:12:42 2025 -0700" }, "message": "net: stmmac: xgmac: Do not enable RX FIFO Overflow interrupts\n\nEnabling RX FIFO Overflow interrupts is counterproductive\nand causes an interrupt storm when RX FIFO overflows.\nDisabling this interrupt has no side effect and eliminates\ninterrupt storms when the RX FIFO overflows.\n\nCommit 8a7cb245cf28 (\"net: stmmac: Do not enable RX FIFO\noverflow interrupts\") disables RX FIFO overflow interrupts\nfor DWMAC4 IP and removes the corresponding handling of\nthis interrupt. This patch is doing the same thing for\nXGMAC IP.\n\nFixes: 2142754f8b9c (\"net: stmmac: Add MAC related callbacks for XGMAC2\")\nSigned-off-by: Rohan G Thomas \u003crohan.g.thomas@altera.com\u003e\nReviewed-by: Matthew Gerlach \u003cmatthew.gerlach@altera.com\u003e\nReviewed-by: Andrew Lunn \u003candrew@lunn.ch\u003e\nLink: https://patch.msgid.link/20250825-xgmac-minor-fixes-v3-1-c225fe4444c0@altera.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "24be0900478dd87b945c7809e343fa07888a24b4", "tree": "68d147edfffb23f9a6e1604661e08814ef9a70e5", "parents": [ "aa125f08cb8e9bc7f08aa7094b78beaeab9bea71", "aca0c31af61e0d5cf1675a0cbd29460b95ae693c" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:29 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:30 2025 -0700" }, "message": "Merge branch \u0027mlx5-misc-fixes-2025-08-25\u0027\n\nMark Bloch says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nmlx5 misc fixes 2025-08-25\n\nThis patchset provides misc bug fixes from the team to the mlx5 core\nand Eth drivers.\n\nv1: https://lore.kernel.org/20250824083944.523858-1-mbloch@nvidia.com\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://patch.msgid.link/20250825143435.598584-1-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "aca0c31af61e0d5cf1675a0cbd29460b95ae693c", "tree": "68d147edfffb23f9a6e1604661e08814ef9a70e5", "parents": [ "d24341740fe48add8a227a753e68b6eedf4b385a" ], "author": { "name": "Alexei Lazar", "email": "alazar@nvidia.com", "time": "Mon Aug 25 17:34:34 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:27 2025 -0700" }, "message": "net/mlx5e: Set local Xoff after FW update\n\nThe local Xoff value is being set before the firmware (FW) update.\nIn case of a failure where the FW is not updated with the new value,\nthere is no fallback to the previous value.\nUpdate the local Xoff value after the FW has been successfully set.\n\nFixes: 0696d60853d5 (\"net/mlx5e: Receive buffer configuration\")\nSigned-off-by: Alexei Lazar \u003calazar@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Dragos Tatulea \u003cdtatulea@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-12-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "d24341740fe48add8a227a753e68b6eedf4b385a", "tree": "a6ec862d5005917677103df6de43f12c999063cb", "parents": [ "ceddedc969f0532b7c62ca971ee50d519d2bc0cb" ], "author": { "name": "Alexei Lazar", "email": "alazar@nvidia.com", "time": "Mon Aug 25 17:34:33 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:27 2025 -0700" }, "message": "net/mlx5e: Update and set Xon/Xoff upon port speed set\n\nXon/Xoff sizes are derived from calculations that include\nthe port speed.\nThese settings need to be updated and applied whenever the\nport speed is changed.\nThe port speed is typically set after the physical link goes down\nand is negotiated as part of the link-up process between the two\nconnected interfaces.\nXon/Xoff parameters being updated at the point where the new\nnegotiated speed is established.\n\nFixes: 0696d60853d5 (\"net/mlx5e: Receive buffer configuration\")\nSigned-off-by: Alexei Lazar \u003calazar@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-11-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "ceddedc969f0532b7c62ca971ee50d519d2bc0cb", "tree": "ad3b599f87eeebbe21588bc232af792ba62501bb", "parents": [ "cf9a8627b9a369ba01d37be6f71b297beb688faa" ], "author": { "name": "Alexei Lazar", "email": "alazar@nvidia.com", "time": "Mon Aug 25 17:34:32 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:27 2025 -0700" }, "message": "net/mlx5e: Update and set Xon/Xoff upon MTU set\n\nXon/Xoff sizes are derived from calculation that include the MTU size.\nSet Xon/Xoff when MTU is set.\nIf Xon/Xoff fails, set the previous MTU.\n\nFixes: 0696d60853d5 (\"net/mlx5e: Receive buffer configuration\")\nSigned-off-by: Alexei Lazar \u003calazar@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-10-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "cf9a8627b9a369ba01d37be6f71b297beb688faa", "tree": "04143ca41766f8bd5c62eaf3dc7980202e989541", "parents": [ "26e42ec7712d392d561964514b1f253b1a96f42d" ], "author": { "name": "Moshe Shemesh", "email": "moshe@nvidia.com", "time": "Mon Aug 25 17:34:31 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:26 2025 -0700" }, "message": "net/mlx5: Prevent flow steering mode changes in switchdev mode\n\nChanging flow steering modes is not allowed when eswitch is in switchdev\nmode. This fix ensures that any steering mode change, including to\nfirmware steering, is correctly blocked while eswitch mode is switchdev.\n\nFixes: e890acd5ff18 (\"net/mlx5: Add devlink flow_steering_mode parameter\")\nSigned-off-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-9-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "26e42ec7712d392d561964514b1f253b1a96f42d", "tree": "b0c1add4d3d903c507bd022323e54ec6fb5375b0", "parents": [ "902a8bc23a24882200f57cadc270e15a2cfaf2bb" ], "author": { "name": "Moshe Shemesh", "email": "moshe@nvidia.com", "time": "Mon Aug 25 17:34:30 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:26 2025 -0700" }, "message": "net/mlx5: Nack sync reset when SFs are present\n\nIf PF (Physical Function) has SFs (Sub-Functions), since the SFs are not\ntaking part in the synchronization flow, sync reset can lead to fatal\nerror on the SFs, as the function will be closed unexpectedly from the\nSF point of view.\n\nAdd a check to prevent sync reset when there are SFs on a PF device\nwhich is not ECPF, as ECPF is teardowned gracefully before reset.\n\nFixes: 92501fa6e421 (\"net/mlx5: Ack on sync_reset_request only if PF can do reset_now\")\nSigned-off-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nReviewed-by: Parav Pandit \u003cparav@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-8-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "902a8bc23a24882200f57cadc270e15a2cfaf2bb", "tree": "55aa0687a829e5ccfa7cd648a3c4382f6cdeee5b", "parents": [ "34cc6a54914f478c93e176450fae6313404f9f74" ], "author": { "name": "Moshe Shemesh", "email": "moshe@nvidia.com", "time": "Mon Aug 25 17:34:29 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:26 2025 -0700" }, "message": "net/mlx5: Fix lockdep assertion on sync reset unload event\n\nFix lockdep assertion triggered during sync reset unload event. When the\nsync reset flow is initiated using the devlink reload fw_activate\noption, the PF already holds the devlink lock while handling unload\nevent. In this case, delegate sync reset unload event handling back to\nthe devlink callback process to avoid double-locking and resolve the\nlockdep warning.\n\nKernel log:\nWARNING: CPU: 9 PID: 1578 at devl_assert_locked+0x31/0x40\n[...]\nCall Trace:\n\u003cTASK\u003e\n mlx5_unload_one_devl_locked+0x2c/0xc0 [mlx5_core]\n mlx5_sync_reset_unload_event+0xaf/0x2f0 [mlx5_core]\n process_one_work+0x222/0x640\n worker_thread+0x199/0x350\n kthread+0x10b/0x230\n ? __pfx_worker_thread+0x10/0x10\n ? __pfx_kthread+0x10/0x10\n ret_from_fork+0x8e/0x100\n ? __pfx_kthread+0x10/0x10\n ret_from_fork_asm+0x1a/0x30\n\u003c/TASK\u003e\n\nFixes: 7a9770f1bfea (\"net/mlx5: Handle sync reset unload event\")\nSigned-off-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-7-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "34cc6a54914f478c93e176450fae6313404f9f74", "tree": "9587e7843a4229d51950781a9aed9b39b431a283", "parents": [ "00a50e4e8974cbf5d6a1dc91cfa5cce4aa7af05a" ], "author": { "name": "Moshe Shemesh", "email": "moshe@nvidia.com", "time": "Mon Aug 25 17:34:28 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:26 2025 -0700" }, "message": "net/mlx5: Reload auxiliary drivers on fw_activate\n\nThe devlink reload fw_activate command performs firmware activation\nfollowed by driver reload, while devlink reload driver_reinit triggers\nonly driver reload. However, the driver reload logic differs between the\ntwo modes, as on driver_reinit mode mlx5 also reloads auxiliary drivers,\nwhile in fw_activate mode the auxiliary drivers are suspended where\napplicable.\n\nAdditionally, following the cited commit, if the device has multiple PFs,\nthe behavior during fw_activate may vary between PFs: one PF may suspend\nauxiliary drivers, while another reloads them.\n\nAlign devlink dev reload fw_activate behavior with devlink dev reload\ndriver_reinit, to reload all auxiliary drivers.\n\nFixes: 72ed5d5624af (\"net/mlx5: Suspend auxiliary devices only in case of PCI device suspend\")\nSigned-off-by: Moshe Shemesh \u003cmoshe@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nReviewed-by: Akiva Goldberger \u003cagoldberger@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-6-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "00a50e4e8974cbf5d6a1dc91cfa5cce4aa7af05a", "tree": "34d8a906690138ab9bb0013db71f1c50ec8668a2", "parents": [ "24b6e53140475b56cadcccd4e82a93aa5bacf1eb" ], "author": { "name": "Lama Kayal", "email": "lkayal@nvidia.com", "time": "Mon Aug 25 17:34:27 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:26 2025 -0700" }, "message": "net/mlx5: HWS, Fix pattern destruction in mlx5hws_pat_get_pattern error path\n\nIn mlx5hws_pat_get_pattern(), when mlx5hws_pat_add_pattern_to_cache()\nfails, the function attempts to clean up the pattern created by\nmlx5hws_cmd_header_modify_pattern_create(). However, it incorrectly\nuses *pattern_id which hasn\u0027t been set yet, instead of the local\nptrn_id variable that contains the actual pattern ID.\n\nThis results in attempting to destroy a pattern using uninitialized\ndata from the output parameter, rather than the valid pattern ID\nreturned by the firmware.\n\nUse ptrn_id instead of *pattern_id in the cleanup path to properly\ndestroy the created pattern.\n\nFixes: aefc15a0fa1c (\"net/mlx5: HWS, added modify header pattern and args handling\")\nSigned-off-by: Lama Kayal \u003clkayal@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-5-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "24b6e53140475b56cadcccd4e82a93aa5bacf1eb", "tree": "0dedfcd31f7ae4419aa495da0e95da55b9d1fe6b", "parents": [ "a630f83592cdad1253523a1b760cfe78fef6cd9c" ], "author": { "name": "Lama Kayal", "email": "lkayal@nvidia.com", "time": "Mon Aug 25 17:34:26 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:26 2025 -0700" }, "message": "net/mlx5: HWS, Fix uninitialized variables in mlx5hws_pat_calc_nop error flow\n\nIn mlx5hws_pat_calc_nop(), src_field and dst_field are passed to\nhws_action_modify_get_target_fields() which should set their values.\nHowever, if an invalid action type is encountered, these variables\nremain uninitialized and are later used to update prev_src_field\nand prev_dst_field.\n\nInitialize both variables to INVALID_FIELD to ensure they have\ndefined values in all code paths.\n\nFixes: 01e035fd0380 (\"net/mlx5: HWS, handle modify header actions dependency\")\nSigned-off-by: Lama Kayal \u003clkayal@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-4-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "a630f83592cdad1253523a1b760cfe78fef6cd9c", "tree": "6f03b2d2ff55c65c51ac2d9dd8e8be66dee17a32", "parents": [ "2c0a959bebdc1ada13cf9a8242f177c5400299e6" ], "author": { "name": "Lama Kayal", "email": "lkayal@nvidia.com", "time": "Mon Aug 25 17:34:25 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:25 2025 -0700" }, "message": "net/mlx5: HWS, Fix memory leak in hws_action_get_shared_stc_nic error flow\n\nWhen an invalid stc_type is provided, the function allocates memory for\nshared_stc but jumps to unlock_and_out without freeing it, causing a\nmemory leak.\n\nFix by jumping to free_shared_stc label instead to ensure proper cleanup.\n\nFixes: 504e536d9010 (\"net/mlx5: HWS, added actions handling\")\nSigned-off-by: Lama Kayal \u003clkayal@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-3-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "2c0a959bebdc1ada13cf9a8242f177c5400299e6", "tree": "d8a88775b3a2c1e47914ade670a9766b8712b00b", "parents": [ "aa125f08cb8e9bc7f08aa7094b78beaeab9bea71" ], "author": { "name": "Lama Kayal", "email": "lkayal@nvidia.com", "time": "Mon Aug 25 17:34:24 2025 +0300" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 18:00:25 2025 -0700" }, "message": "net/mlx5: HWS, Fix memory leak in hws_pool_buddy_init error path\n\nIn the error path of hws_pool_buddy_init(), the buddy allocator cleanup\ndoesn\u0027t free the allocator structure itself, causing a memory leak.\n\nAdd the missing kfree() to properly release all allocated memory.\n\nFixes: c61afff94373 (\"net/mlx5: HWS, added memory management handling\")\nSigned-off-by: Lama Kayal \u003clkayal@nvidia.com\u003e\nReviewed-by: Tariq Toukan \u003ctariqt@nvidia.com\u003e\nSigned-off-by: Mark Bloch \u003cmbloch@nvidia.com\u003e\nLink: https://patch.msgid.link/20250825143435.598584-2-mbloch@nvidia.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "aa125f08cb8e9bc7f08aa7094b78beaeab9bea71", "tree": "9cc524119390dedf2d688a6c8754de4c15af4d48", "parents": [ "e69880cbf623dc6283ae798aa5b8fb930063591c", "ed913b343dcf9f623e7436fa1a153c89b22d109b" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:50:33 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:50:34 2025 -0700" }, "message": "Merge branch \u0027100GbE\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue\n\nTony Nguyen says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nIntel Wired LAN Driver Updates 2025-08-25 (ice, ixgbe)\n\nFor ice:\nEmil adds a check to ensure auxiliary device was created before tear\ndown to prevent NULL a pointer dereference.\n\nJake reworks flow for failed Tx scheduler configuration to allow for\nproper recovery and operation. He also adjusts ice_adapter index for\nE825C devices as use of DSN is incompatible with this device.\n\nMichal corrects tracking of buffer allocation failure in\nice_clean_rx_irq().\n\nFor ixgbe:\nJedrzej adds __packed attribute to ixgbe_orom_civd_info to compatibility\nwith device OROM data.\n\n* \u0027100GbE\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:\n ixgbe: fix ixgbe_orom_civd_info struct layout\n ice: fix incorrect counter for buffer allocation failures\n ice: use fixed adapter index for E825C embedded devices\n ice: don\u0027t leave device non-functional if Tx scheduler config fails\n ice: fix NULL pointer dereference in ice_unplug_aux_dev() on reset\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://patch.msgid.link/20250825215019.3442873-1-anthony.l.nguyen@intel.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "e69880cbf623dc6283ae798aa5b8fb930063591c", "tree": "18aab7b1b5009769966a93e1f755801437b4affe", "parents": [ "d9b0ca1334d8a9a03bef45e95825564c56ca3367", "b4fc8faacfea2538184a1dbd616ae9447a361f3d" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:48:37 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:48:38 2025 -0700" }, "message": "Merge branch \u0027bnxt_en-3-bug-fixes\u0027\n\nMichael Chan says:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nbnxt_en: 3 bug fixes\n\nThe first one fixes a memory corruption issue that can happen when\nFW resources change during ifdown with TCs created. The next two\nfix FW resource reservation logic for TX rings and stats context.\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\n\nLink: https://patch.msgid.link/20250825175927.459987-1-michael.chan@broadcom.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "b4fc8faacfea2538184a1dbd616ae9447a361f3d", "tree": "18aab7b1b5009769966a93e1f755801437b4affe", "parents": [ "1ee581c24dfdcbc6de25aac95a48c1f08e9a542c" ], "author": { "name": "Michael Chan", "email": "michael.chan@broadcom.com", "time": "Mon Aug 25 10:59:27 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:48:35 2025 -0700" }, "message": "bnxt_en: Fix stats context reservation logic\n\nThe HW resource reservation logic allows the L2 driver to use the\nRoCE resources if the RoCE driver is not registered. When calculating\nthe stats contexts available for L2, we should not blindly subtract\nthe stats contexts reserved for RoCE unless the RoCE driver is\nregistered. This bug may cause the L2 rings to be less than the\nnumber requested when we are close to running out of stats contexts.\n\nFixes: 2e4592dc9bee (\"bnxt_en: Change MSIX/NQs allocation policy\")\nReviewed-by: Kalesh AP \u003ckalesh-anakkur.purayil@broadcom.com\u003e\nReviewed-by: Somnath Kotur \u003csomnath.kotur@broadcom.com\u003e\nSigned-off-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nLink: https://patch.msgid.link/20250825175927.459987-4-michael.chan@broadcom.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "1ee581c24dfdcbc6de25aac95a48c1f08e9a542c", "tree": "82aba09ec7e6517d2cdb98cdad3ce2cf9bd10557", "parents": [ "2747328ba2714f1a7454208dbbc1dc0631990b4a" ], "author": { "name": "Michael Chan", "email": "michael.chan@broadcom.com", "time": "Mon Aug 25 10:59:26 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:48:35 2025 -0700" }, "message": "bnxt_en: Adjust TX rings if reservation is less than requested\n\nBefore we accept an ethtool request to increase a resource (such as\nrings), we call the FW to check that the requested resource is likely\navailable first before we commit. But it is still possible that\nthe actual reservation or allocation can fail. The existing code\nis missing the logic to adjust the TX rings in case the reserved\nTX rings are less than requested. Add a warning message (a similar\nmessage for RX rings already exists) and add the logic to adjust\nthe TX rings. Without this fix, the number of TX rings reported\nto the stack can exceed the actual TX rings and ethtool -l will\nreport more than the actual TX rings.\n\nFixes: 674f50a5b026 (\"bnxt_en: Implement new method to reserve rings.\")\nReviewed-by: Kalesh AP \u003ckalesh-anakkur.purayil@broadcom.com\u003e\nReviewed-by: Somnath Kotur \u003csomnath.kotur@broadcom.com\u003e\nSigned-off-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nLink: https://patch.msgid.link/20250825175927.459987-3-michael.chan@broadcom.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "2747328ba2714f1a7454208dbbc1dc0631990b4a", "tree": "3f9a8a84a55d1cedceeb78c50d57dcd1497fd152", "parents": [ "d9b0ca1334d8a9a03bef45e95825564c56ca3367" ], "author": { "name": "Sreekanth Reddy", "email": "sreekanth.reddy@broadcom.com", "time": "Mon Aug 25 10:59:25 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:48:35 2025 -0700" }, "message": "bnxt_en: Fix memory corruption when FW resources change during ifdown\n\nbnxt_set_dflt_rings() assumes that it is always called before any TC has\nbeen created. So it doesn\u0027t take bp-\u003enum_tc into account and assumes\nthat it is always 0 or 1.\n\nIn the FW resource or capability change scenario, the FW will return\nflags in bnxt_hwrm_if_change() that will cause the driver to\nreinitialize and call bnxt_cancel_reservations(). This will lead to\nbnxt_init_dflt_ring_mode() calling bnxt_set_dflt_rings() and bp-\u003enum_tc\nmay be greater than 1. This will cause bp-\u003etx_ring[] to be sized too\nsmall and cause memory corruption in bnxt_alloc_cp_rings().\n\nFix it by properly scaling the TX rings by bp-\u003enum_tc in the code\npaths mentioned above. Add 2 helper functions to determine\nbp-\u003etx_nr_rings and bp-\u003etx_nr_rings_per_tc.\n\nFixes: ec5d31e3c15d (\"bnxt_en: Handle firmware reset status during IF_UP.\")\nReviewed-by: Kalesh AP \u003ckalesh-anakkur.purayil@broadcom.com\u003e\nReviewed-by: Andy Gospodarek \u003candrew.gospodarek@broadcom.com\u003e\nSigned-off-by: Sreekanth Reddy \u003csreekanth.reddy@broadcom.com\u003e\nSigned-off-by: Michael Chan \u003cmichael.chan@broadcom.com\u003e\nLink: https://patch.msgid.link/20250825175927.459987-2-michael.chan@broadcom.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "d9b0ca1334d8a9a03bef45e95825564c56ca3367", "tree": "4a4acf68d3864a0d0b98e91d51cbe8229efb3bd7", "parents": [ "16c8a3a67ec799fc731919e3e51be9af6cdf541d" ], "author": { "name": "Boon Khai Ng", "email": "boon.khai.ng@altera.com", "time": "Mon Aug 25 15:13:21 2025 +0800" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:39:25 2025 -0700" }, "message": "MAINTAINERS: Update maintainer information for Altera Triple Speed Ethernet Driver\n\nThe previous maintainer, Joyce Ooi, is no longer with the company,\nand her email is no longer reachable. As a result, the maintainer\ninformation for the Altera Triple Speed Ethernet Driver has been updated.\n\nChanges:\n- Replaced Joyce Ooi\u0027s email with Boon Khai Ng\u0027s email address.\n- Kept the component\u0027s status as \"Maintained\".\n\nSigned-off-by: Boon Khai Ng \u003cboon.khai.ng@altera.com\u003e\nLink: https://patch.msgid.link/20250825071321.30131-1-boon.khai.ng@altera.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "16c8a3a67ec799fc731919e3e51be9af6cdf541d", "tree": "b9b3279f50e65eb77a7006a554ad8206ebbc1560", "parents": [ "26c1f55f7ec8d1a4bde8c50e4ee04e3c8c6b27e8" ], "author": { "name": "Sean Anderson", "email": "sean.anderson@linux.dev", "time": "Mon Aug 25 13:21:34 2025 -0400" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:36:28 2025 -0700" }, "message": "net: macb: Fix offset error in gem_update_stats\n\nhw_stats now has only one variable for tx_octets/rx_octets, so we should\nonly increment p once, not twice. This would cause the statistics to be\nreported under the wrong categories in `ethtool -S --all-groups` (which\nuses hw_stats) but not `ethtool -S` (which uses ethtool_stats).\n\nSigned-off-by: Sean Anderson \u003csean.anderson@linux.dev\u003e\nFixes: f6af690a295a (\"net: cadence: macb: Report standard stats\")\nLink: https://patch.msgid.link/20250825172134.681861-1-sean.anderson@linux.dev\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "26c1f55f7ec8d1a4bde8c50e4ee04e3c8c6b27e8", "tree": "89f354ce79dd2d8375a0b97554e6b28de79a8623", "parents": [ "882e57cbc7204662f6c5672d5b04336c1d790b03" ], "author": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Mon Aug 25 08:57:53 2025 -0700" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:36:01 2025 -0700" }, "message": "MAINTAINERS: retire Boris from TLS maintainers\n\nThere\u0027s a steady stream of TLS changes and bugs. We need active\nmaintainers in this area, and Boris hasn\u0027t been participating\nmuch in upstream work. Move him to CREDITS. While at it also\nadd Dave Watson there who was the author of the initial SW\nimplementation, AFAIU.\n\nLink: https://patch.msgid.link/20250825155753.2178045-1-kuba@kernel.org\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "882e57cbc7204662f6c5672d5b04336c1d790b03", "tree": "ec2aa2cd87d17ed7d6cf56d816d337c40c36a702", "parents": [ "e81a7f65288c7e2cfb7e7890f648e099fd885ab3" ], "author": { "name": "Horatiu Vultur", "email": "horatiu.vultur@microchip.com", "time": "Mon Aug 25 08:55:43 2025 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:10:51 2025 -0700" }, "message": "phy: mscc: Fix when PTP clock is register and unregister\n\nIt looks like that every time when the interface was set down and up the\ndriver was creating a new ptp clock. On top of this the function\nptp_clock_unregister was never called.\nTherefore fix this by calling ptp_clock_register and initialize the\nmii_ts struct inside the probe function and call ptp_clock_unregister when\ndriver is removed.\n\nFixes: 7d272e63e0979d (\"net: phy: mscc: timestamping and PHC support\")\nSigned-off-by: Horatiu Vultur \u003choratiu.vultur@microchip.com\u003e\nReviewed-by: Vadim Fedorenko \u003cvadim.fedorenko@linux.dev\u003e\nReviewed-by: Vladimir Oltean \u003cvladimir.oltean@nxp.com\u003e\nLink: https://patch.msgid.link/20250825065543.2916334-1-horatiu.vultur@microchip.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "e81a7f65288c7e2cfb7e7890f648e099fd885ab3", "tree": "1e4aeb738cf2a2c67a5a3f916113d4a51ed0ee5d", "parents": [ "007a5ffadc4fd51739527f1503b7cf048f31c413" ], "author": { "name": "Fabio Porcedda", "email": "fabio.porcedda@gmail.com", "time": "Fri Aug 22 11:13:24 2025 +0200" }, "committer": { "name": "Jakub Kicinski", "email": "kuba@kernel.org", "time": "Tue Aug 26 17:08:42 2025 -0700" }, "message": "net: usb: qmi_wwan: add Telit Cinterion LE910C4-WWX new compositions\n\nAdd the following Telit Cinterion LE910C4-WWX new compositions:\n\n0x1034: tty (AT) + tty (AT) + rmnet\nT: Bus\u003d01 Lev\u003d01 Prnt\u003d01 Port\u003d00 Cnt\u003d01 Dev#\u003d 8 Spd\u003d480 MxCh\u003d 0\nD: Ver\u003d 2.00 Cls\u003d00(\u003eifc ) Sub\u003d00 Prot\u003d00 MxPS\u003d64 #Cfgs\u003d 1\nP: Vendor\u003d1bc7 ProdID\u003d1034 Rev\u003d00.00\nS: Manufacturer\u003dTelit\nS: Product\u003dLE910C4-WWX\nS: SerialNumber\u003d93f617e7\nC: #Ifs\u003d 3 Cfg#\u003d 1 Atr\u003de0 MxPwr\u003d500mA\nI: If#\u003d 0 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dff Prot\u003dff Driver\u003doption\nE: Ad\u003d01(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d81(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d82(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 1 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dfe Prot\u003dff Driver\u003doption\nE: Ad\u003d02(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d83(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d84(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 2 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dff Prot\u003dff Driver\u003dqmi_wwan\nE: Ad\u003d03(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d85(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d86(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\n\n0x1037: tty (diag) + tty (Telit custom) + tty (AT) + tty (AT) + rmnet\nT: Bus\u003d01 Lev\u003d01 Prnt\u003d01 Port\u003d00 Cnt\u003d01 Dev#\u003d 15 Spd\u003d480 MxCh\u003d 0\nD: Ver\u003d 2.00 Cls\u003d00(\u003eifc ) Sub\u003d00 Prot\u003d00 MxPS\u003d64 #Cfgs\u003d 1\nP: Vendor\u003d1bc7 ProdID\u003d1037 Rev\u003d00.00\nS: Manufacturer\u003dTelit\nS: Product\u003dLE910C4-WWX\nS: SerialNumber\u003d93f617e7\nC: #Ifs\u003d 5 Cfg#\u003d 1 Atr\u003de0 MxPwr\u003d500mA\nI: If#\u003d 0 Alt\u003d 0 #EPs\u003d 2 Cls\u003dff(vend.) Sub\u003dff Prot\u003d30 Driver\u003doption\nE: Ad\u003d01(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d81(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 1 Alt\u003d 0 #EPs\u003d 2 Cls\u003dff(vend.) Sub\u003dff Prot\u003dff Driver\u003doption\nE: Ad\u003d02(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d82(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 2 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dff Prot\u003dff Driver\u003doption\nE: Ad\u003d03(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d83(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d84(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 3 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dfe Prot\u003dff Driver\u003doption\nE: Ad\u003d04(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d85(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d86(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 4 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dff Prot\u003dff Driver\u003dqmi_wwan\nE: Ad\u003d05(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d87(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d88(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\n\n0x1038: tty (Telit custom) + tty (AT) + tty (AT) + rmnet\nT: Bus\u003d01 Lev\u003d01 Prnt\u003d01 Port\u003d00 Cnt\u003d01 Dev#\u003d 9 Spd\u003d480 MxCh\u003d 0\nD: Ver\u003d 2.00 Cls\u003d00(\u003eifc ) Sub\u003d00 Prot\u003d00 MxPS\u003d64 #Cfgs\u003d 1\nP: Vendor\u003d1bc7 ProdID\u003d1038 Rev\u003d00.00\nS: Manufacturer\u003dTelit\nS: Product\u003dLE910C4-WWX\nS: SerialNumber\u003d93f617e7\nC: #Ifs\u003d 4 Cfg#\u003d 1 Atr\u003de0 MxPwr\u003d500mA\nI: If#\u003d 0 Alt\u003d 0 #EPs\u003d 2 Cls\u003dff(vend.) Sub\u003dff Prot\u003dff Driver\u003doption\nE: Ad\u003d01(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d81(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 1 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dff Prot\u003dff Driver\u003doption\nE: Ad\u003d02(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d82(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d83(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 2 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dfe Prot\u003dff Driver\u003doption\nE: Ad\u003d03(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d84(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d85(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nI: If#\u003d 3 Alt\u003d 0 #EPs\u003d 3 Cls\u003dff(vend.) Sub\u003dff Prot\u003dff Driver\u003dqmi_wwan\nE: Ad\u003d04(O) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\nE: Ad\u003d86(I) Atr\u003d03(Int.) MxPS\u003d 64 Ivl\u003d2ms\nE: Ad\u003d87(I) Atr\u003d02(Bulk) MxPS\u003d 512 Ivl\u003d0ms\n\nCc: stable@vger.kernel.org\nSigned-off-by: Fabio Porcedda \u003cfabio.porcedda@gmail.com\u003e\nLink: https://patch.msgid.link/20250822091324.39558-1-Fabio.Porcedda@telit.com\nSigned-off-by: Jakub Kicinski \u003ckuba@kernel.org\u003e\n" }, { "commit": "e3ef9445cd9d90e43de0bd3cd55d437773dfd139", "tree": "19a5d3a4ea7e4bd19ff4f7d4dfca1122457ff44d", "parents": [ "198f36f902ec7e99b645382505f74b87a4523ed9" ], "author": { "name": "Nilay Shroff", "email": "nilay@linux.ibm.com", "time": "Tue Aug 26 22:00:32 2025 +0530" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Tue Aug 26 10:34:08 2025 -0600" }, "message": "block: validate QoS before calling __rq_qos_done_bio()\n\nIf a bio has BIO_QOS_xxx set, it doesn\u0027t guarantee that q-\u003erq_qos is\nalso present at-least for stacked block devices. For instance, in case\nof NVMe when multipath is enabled, the bottom device may have QoS\nenabled but top device doesn\u0027t. So always validate QoS is enabled and\nq-\u003erq_qos is present before calling __rq_qos_done_bio().\n\nFixes: 370ac285f23a (\"block: avoid cpu_hotplug_lock depedency on freeze_lock\")\nReported-by: Venkat Rao Bagalkote \u003cvenkat88@linux.ibm.com\u003e\nCloses: https://lore.kernel.org/all/3a07b752-06a4-4eee-b302-f4669feb859d@linux.ibm.com/\nSigned-off-by: Nilay Shroff \u003cnilay@linux.ibm.com\u003e\nLink: https://lore.kernel.org/r/20250826163128.1952394-1-nilay@linux.ibm.com\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "198f36f902ec7e99b645382505f74b87a4523ed9", "tree": "9ba0766d178becba985618203cf5c9dbbfea6cc3", "parents": [ "d14469ed7c00314fe8957b2841bda329e4eaf4ab" ], "author": { "name": "Bart Van Assche", "email": "bvanassche@acm.org", "time": "Mon Aug 25 11:27:19 2025 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@kernel.dk", "time": "Tue Aug 26 08:27:24 2025 -0600" }, "message": "blk-zoned: Fix a lockdep complaint about recursive locking\n\nIf preparing a write bio fails then blk_zone_wplug_bio_work() calls\nbio_endio() with zwplug-\u003elock held. If a device mapper driver is stacked\non top of the zoned block device then this results in nested locking of\nzwplug-\u003elock. The resulting lockdep complaint is a false positive\nbecause this is nested locking and not recursive locking. Suppress this\nfalse positive by calling blk_zone_wplug_bio_io_error() without holding\nzwplug-\u003elock. This is safe because no code in\nblk_zone_wplug_bio_io_error() depends on zwplug-\u003elock being held. This\npatch suppresses the following lockdep complaint:\n\nWARNING: possible recursive locking detected\n--------------------------------------------\nkworker/3:0H/46 is trying to acquire lock:\nffffff882968b830 (\u0026zwplug-\u003elock){-...}-{2:2}, at: blk_zone_write_plug_bio_endio+0x64/0x1f0\n\nbut task is already holding lock:\nffffff88315bc230 (\u0026zwplug-\u003elock){-...}-{2:2}, at: blk_zone_wplug_bio_work+0x8c/0x48c\n\nother info that might help us debug this:\n Possible unsafe locking scenario:\n\n CPU0\n ----\n lock(\u0026zwplug-\u003elock);\n lock(\u0026zwplug-\u003elock);\n\n *** DEADLOCK ***\n\n May be due to missing lock nesting notation\n\n3 locks held by kworker/3:0H/46:\n #0: ffffff8809486758 ((wq_completion)sdd_zwplugs){+.+.}-{0:0}, at: process_one_work+0x1bc/0x65c\n #1: ffffffc085de3d70 ((work_completion)(\u0026zwplug-\u003ebio_work)){+.+.}-{0:0}, at: process_one_work+0x1e4/0x65c\n #2: ffffff88315bc230 (\u0026zwplug-\u003elock){-...}-{2:2}, at: blk_zone_wplug_bio_work+0x8c/0x48c\n\nstack backtrace:\nCPU: 3 UID: 0 PID: 46 Comm: kworker/3:0H Tainted: G W OE 6.12.38-android16-5-maybe-dirty-4k #1 8b362b6f76e3645a58cd27d86982bce10d150025\nTainted: [W]\u003dWARN, [O]\u003dOOT_MODULE, [E]\u003dUNSIGNED_MODULE\nHardware name: Spacecraft board based on MALIBU (DT)\nWorkqueue: sdd_zwplugs blk_zone_wplug_bio_work\nCall trace:\n dump_backtrace+0xfc/0x17c\n show_stack+0x18/0x28\n dump_stack_lvl+0x40/0xa0\n dump_stack+0x18/0x24\n print_deadlock_bug+0x38c/0x398\n __lock_acquire+0x13e8/0x2e1c\n lock_acquire+0x134/0x2b4\n _raw_spin_lock_irqsave+0x5c/0x80\n blk_zone_write_plug_bio_endio+0x64/0x1f0\n bio_endio+0x9c/0x240\n __dm_io_complete+0x214/0x260\n clone_endio+0xe8/0x214\n bio_endio+0x218/0x240\n blk_zone_wplug_bio_work+0x204/0x48c\n process_one_work+0x26c/0x65c\n worker_thread+0x33c/0x498\n kthread+0x110/0x134\n ret_from_fork+0x10/0x20\n\nCc: stable@vger.kernel.org\nCc: Damien Le Moal \u003cdlemoal@kernel.org\u003e\nCc: Christoph Hellwig \u003chch@lst.de\u003e\nFixes: dd291d77cc90 (\"block: Introduce zone write plugging\")\nSigned-off-by: Bart Van Assche \u003cbvanassche@acm.org\u003e\nReviewed-by: Damien Le Moal \u003cdlemoal@kernel.org\u003e\nLink: https://lore.kernel.org/r/20250825182720.1697203-1-bvanassche@acm.org\nSigned-off-by: Jens Axboe \u003caxboe@kernel.dk\u003e\n" }, { "commit": "75671d90fde8c78e940e15a1366a50ece56c6b69", "tree": "0f4167d38d72cd20289878118b165ca2d4a28ec5", "parents": [ "16ca06aa2c2218cb21907c0c45a746958c944def" ], "author": { "name": "Carlos Llamas", "email": "cmllamas@google.com", "time": "Mon Aug 25 15:57:42 2025 +0000" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Tue Aug 26 10:12:11 2025 -0400" }, "message": "drm/xe: switch to local xbasename() helper\n\nCommit b0a2ee5567ab (\"drm/xe: prepare xe_gen_wa_oob to be multi-use\")\nintroduced a call to basename(). The GNU version of this function is not\nportable and fails to build with alternative libc implementations like\nmusl or bionic. This causes the following build error:\n\n drivers/gpu/drm/xe/xe_gen_wa_oob.c:130:12: error: assignment to ‘const char *’ from ‘int’ makes pointer from integer without a cast [-Wint-conversion]\n 130 | fn \u003d basename(fn);\n | ^\n\nWhile a POSIX version of basename() could be used, it would require a\nseparate header plus the behavior differs from GNU version in that it\nmight modify its argument. Not great.\n\nInstead, implement a local xbasename() helper based on strrchr() that\nprovides the same functionality and avoids portability issues.\n\nFixes: b0a2ee5567ab (\"drm/xe: prepare xe_gen_wa_oob to be multi-use\")\nSuggested-by: Lucas De Marchi \u003clucas.demarchi@intel.com\u003e\nReviewed-by: Tiffany Yang \u003cynaffit@google.com\u003e\nSigned-off-by: Carlos Llamas \u003ccmllamas@google.com\u003e\nReviewed-by: Lucas De Marchi \u003clucas.demarchi@intel.com\u003e\nLink: https://lore.kernel.org/r/20250825155743.1132433-1-cmllamas@google.com\nSigned-off-by: Lucas De Marchi \u003clucas.demarchi@intel.com\u003e\n(cherry picked from commit 41be792f5baaf90d744a9a9e82994ce560ca9582)\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\n" }, { "commit": "16ca06aa2c2218cb21907c0c45a746958c944def", "tree": "0f419f6affe4c2f4a8ee0bcb63ca9a1ffffc56fb", "parents": [ "2b55ddf36229e0278c956215784ab1feeff510aa" ], "author": { "name": "Matthew Brost", "email": "matthew.brost@intel.com", "time": "Mon Aug 25 08:28:41 2025 -0700" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Tue Aug 26 10:12:11 2025 -0400" }, "message": "drm/xe: Don\u0027t trigger rebind on initial dma-buf validation\n\nOn the first validate of an imported dma-buf (initial bind), the device\nhas no GPU mappings, so a rebind is unnecessary. Rebinding here is\nharmful in multi-GPU setups and for VMs using preempt-fence mode, as it\nwould evict in-flight GPU work.\n\nv2:\n - Drop dma_buf_validated, check for XE_PL_SYSTEM (Thomas)\n\nFixes: dd08ebf6c352 (\"drm/xe: Introduce a new DRM driver for Intel GPUs\")\nSigned-off-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nReviewed-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\nLink: https://lore.kernel.org/r/20250825152841.3837378-1-matthew.brost@intel.com\n(cherry picked from commit ffdf968762e4fb3cdae54e811ec3525e67440a60)\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\n" }, { "commit": "2b55ddf36229e0278c956215784ab1feeff510aa", "tree": "68324465471855325f7073d40fd19154c38733cf", "parents": [ "7551865cd12af2dc47e5a174eebcfb0b94b5449b" ], "author": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Aug 21 16:30:45 2025 +0200" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Tue Aug 26 10:12:11 2025 -0400" }, "message": "drm/xe/vm: Clear the scratch_pt pointer on error\n\nAvoid triggering a dereference of an error pointer on cleanup in\nxe_vm_free_scratch() by clearing any scratch_pt error pointer.\n\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\nFixes: 06951c2ee72d (\"drm/xe: Use NULL PTEs as scratch PTEs\")\nCc: Brian Welty \u003cbrian.welty@intel.com\u003e\nCc: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\nCc: Lucas De Marchi \u003clucas.demarchi@intel.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e # v6.8+\nReviewed-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nLink: https://lore.kernel.org/r/20250821143045.106005-4-thomas.hellstrom@linux.intel.com\n(cherry picked from commit 358ee50ab565f3c8ea32480e9d03127a81ba32f8)\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\n" }, { "commit": "7551865cd12af2dc47e5a174eebcfb0b94b5449b", "tree": "1807d2efbf2240c0a5c28c5ee66d0e54eb53019c", "parents": [ "04e1f683cd28dc9407b238543871a6e09a570dc0" ], "author": { "name": "Thomas Hellström", "email": "thomas.hellstrom@linux.intel.com", "time": "Thu Aug 21 16:30:43 2025 +0200" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Tue Aug 26 10:12:11 2025 -0400" }, "message": "drm/xe/vm: Don\u0027t pin the vm_resv during validation\n\nThe pinning has the odd side-effect that unlocking *any* resv\nduring validation triggers an \"unlocking pinned lock\" warning.\n\nCc: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nFixes: 5cc3325584c4 (\"drm/xe: Rework eviction rejection of bound external bos\")\nSigned-off-by: Thomas Hellström \u003cthomas.hellstrom@linux.intel.com\u003e\nReviewed-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nLink: https://lore.kernel.org/r/20250821143045.106005-2-thomas.hellstrom@linux.intel.com\n(cherry picked from commit 0a51bf3e54dd8b77e6f1febbbb66def0660862d2)\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\n" }, { "commit": "04e1f683cd28dc9407b238543871a6e09a570dc0", "tree": "814b445ebeb01ac21b4a200cf0592453046a6e26", "parents": [ "1b237f190eb3d36f52dffe07a40b5eb210280e00" ], "author": { "name": "Zbigniew Kempczyński", "email": "zbigniew.kempczynski@intel.com", "time": "Wed Aug 20 10:39:04 2025 +0200" }, "committer": { "name": "Rodrigo Vivi", "email": "rodrigo.vivi@intel.com", "time": "Tue Aug 26 10:12:11 2025 -0400" }, "message": "drm/xe/xe_sync: avoid race during ufence signaling\n\nMarking ufence as signalled after copy_to_user() is too late.\nWorker thread which signals ufence by memory write might be raced\nwith another userspace vm-bind call. In map/unmap scenario unmap\nmay still see ufence is not signalled causing -EBUSY. Change the\norder of marking / write to user-fence fixes this issue.\n\nFixes: 977e5b82e090 (\"drm/xe: Expose user fence from xe_sync_entry\")\nLink: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/5536\nSigned-off-by: Zbigniew Kempczyński \u003czbigniew.kempczynski@intel.com\u003e\nCc: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nCc: Matthew Auld \u003cmatthew.auld@intel.com\u003e\nReviewed-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nSigned-off-by: Matthew Brost \u003cmatthew.brost@intel.com\u003e\nLink: https://lore.kernel.org/r/20250820083903.2109891-2-zbigniew.kempczynski@intel.com\n(cherry picked from commit 8ae04fe9ffc93d6bc3bc63ac08375427d69cee06)\nSigned-off-by: Rodrigo Vivi \u003crodrigo.vivi@intel.com\u003e\n" }, { "commit": "e246518aa24f1460902725e97d0abf574aec6ade", "tree": "626fbde3d3a2a5354881b985cae383be153f26f9", "parents": [ "1b237f190eb3d36f52dffe07a40b5eb210280e00" ], "author": { "name": "Johannes Berg", "email": "johannes.berg@intel.com", "time": "Tue Aug 26 13:43:47 2025 +0200" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Tue Aug 26 15:27:04 2025 +0200" }, "message": "PM: sleep: annotate RCU list iterations\n\nThese iterations require the read lock, otherwise RCU\nlockdep will splat:\n\n\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\nWARNING: suspicious RCU usage\n6.17.0-rc3-00014-g31419c045d64 #6 Tainted: G O\n-----------------------------\ndrivers/base/power/main.c:1333 RCU-list traversed in non-reader section!!\n\nother info that might help us debug this:\n\nrcu_scheduler_active \u003d 2, debug_locks \u003d 1\n5 locks held by rtcwake/547:\n #0: 00000000643ab418 (sb_writers#6){.+.+}-{0:0}, at: file_start_write+0x2b/0x3a\n #1: 0000000067a0ca88 (\u0026of-\u003emutex#2){+.+.}-{4:4}, at: kernfs_fop_write_iter+0x181/0x24b\n #2: 00000000631eac40 (kn-\u003eactive#3){.+.+}-{0:0}, at: kernfs_fop_write_iter+0x191/0x24b\n #3: 00000000609a1308 (system_transition_mutex){+.+.}-{4:4}, at: pm_suspend+0xaf/0x30b\n #4: 0000000060c0fdb0 (device_links_srcu){.+.+}-{0:0}, at: device_links_read_lock+0x75/0x98\n\nstack backtrace:\nCPU: 0 UID: 0 PID: 547 Comm: rtcwake Tainted: G O 6.17.0-rc3-00014-g31419c045d64 #6 VOLUNTARY\nTainted: [O]\u003dOOT_MODULE\nStack:\n 223721b3a80 6089eac6 00000001 00000001\n ffffff00 6089eac6 00000535 6086e528\n 721b3ac0 6003c294 00000000 60031fc0\nCall Trace:\n [\u003c600407ed\u003e] show_stack+0x10e/0x127\n [\u003c6003c294\u003e] dump_stack_lvl+0x77/0xc6\n [\u003c6003c2fd\u003e] dump_stack+0x1a/0x20\n [\u003c600bc2f8\u003e] lockdep_rcu_suspicious+0x116/0x13e\n [\u003c603d8ea1\u003e] dpm_async_suspend_superior+0x117/0x17e\n [\u003c603d980f\u003e] device_suspend+0x528/0x541\n [\u003c603da24b\u003e] dpm_suspend+0x1a2/0x267\n [\u003c603da837\u003e] dpm_suspend_start+0x5d/0x72\n [\u003c600ca0c9\u003e] suspend_devices_and_enter+0xab/0x736\n [...]\n\nAdd the fourth argument to the iteration to annotate\nthis and avoid the splat.\n\nFixes: 06799631d522 (\"PM: sleep: Make async suspend handle suppliers like parents\")\nFixes: ed18738fff02 (\"PM: sleep: Make async resume handle consumers like children\")\nSigned-off-by: Johannes Berg \u003cjohannes.berg@intel.com\u003e\nLink: https://patch.msgid.link/20250826134348.aba79f6e6299.I9ecf55da46ccf33778f2c018a82e1819d815b348@changeid\nSigned-off-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\n" }, { "commit": "134ed1093907a79e5d98087513f13fd7652c4df9", "tree": "61eacab6d7a70a2c1c85927d9f74fa67437f10dc", "parents": [ "01a3044af5d910e1f8b0bee53b2e0eccee8f9a5c" ], "author": { "name": "Jan Kiszka", "email": "jan.kiszka@siemens.com", "time": "Mon Aug 25 18:07:13 2025 +0200" }, "committer": { "name": "Ard Biesheuvel", "email": "ardb@kernel.org", "time": "Tue Aug 26 14:49:42 2025 +0200" }, "message": "efi: stmm: Drop unneeded null pointer check\n\nThe API documenation of setup_mm_hdr does not mention that dptr can be\nNULL, this is a local function, and no caller passes NULL. So drop the\nunneeded check.\n\nSigned-off-by: Jan Kiszka \u003cjan.kiszka@siemens.com\u003e\nReviewed-by: Ilias Apalodimas \u003cilias.apalodimas@linaro.org\u003e\nAcked-by: Sumit Garg \u003csumit.garg@oss.qualcomm.com\u003e\nSigned-off-by: Ard Biesheuvel \u003cardb@kernel.org\u003e\n" }, { "commit": "01a3044af5d910e1f8b0bee53b2e0eccee8f9a5c", "tree": "86da47e2c1672e131b3b0f5015dc9c9b5937e633", "parents": [ "80c6c1048625712200ab9cdc665d792b85594e2c" ], "author": { "name": "Jan Kiszka", "email": "jan.kiszka@siemens.com", "time": "Mon Aug 25 18:07:12 2025 +0200" }, "committer": { "name": "Ard Biesheuvel", "email": "ardb@kernel.org", "time": "Tue Aug 26 14:49:42 2025 +0200" }, "message": "efi: stmm: Drop unused EFI error from setup_mm_hdr arguments\n\nNo caller ever evaluates what we return in \u0027ret\u0027. They only use the\nreturn code of the function.\n\nSigned-off-by: Jan Kiszka \u003cjan.kiszka@siemens.com\u003e\nReviewed-by: Ilias Apalodimas \u003cilias.apalodimas@linaro.org\u003e\nAcked-by: Sumit Garg \u003csumit.garg@oss.qualcomm.com\u003e\nSigned-off-by: Ard Biesheuvel \u003cardb@kernel.org\u003e\n" }, { "commit": "80c6c1048625712200ab9cdc665d792b85594e2c", "tree": "3f1304f3744ce15e79a2e302ed04a40e7c294b01", "parents": [ "c5e81e672699e0c5557b2b755cc8f7a69aa92bff" ], "author": { "name": "Jan Kiszka", "email": "jan.kiszka@siemens.com", "time": "Mon Aug 25 18:07:11 2025 +0200" }, "committer": { "name": "Ard Biesheuvel", "email": "ardb@kernel.org", "time": "Tue Aug 26 14:49:42 2025 +0200" }, "message": "efi: stmm: Do not return EFI_OUT_OF_RESOURCES on internal errors\n\nWhen we are low on memory or when the internal API is violated, we\ncannot return EFI_OUT_OF_RESOURCES. According to the UEFI standard, that\nerror code is either related to persistent storage used for the variable\nor even not foreseen as possible error (GetVariable e.g.). Use the not\nfully accurate but compliant error code EFI_DEVICE_ERROR in those cases.\n\nSigned-off-by: Jan Kiszka \u003cjan.kiszka@siemens.com\u003e\nReviewed-by: Ilias Apalodimas \u003cilias.apalodimas@linaro.org\u003e\nAcked-by: Sumit Garg \u003csumit.garg@oss.qualcomm.com\u003e\nSigned-off-by: Ard Biesheuvel \u003cardb@kernel.org\u003e\n" }, { "commit": "c5e81e672699e0c5557b2b755cc8f7a69aa92bff", "tree": "c380524326d1a59c5e82286155473a958c1b9266", "parents": [ "8f5ae30d69d7543eee0d70083daf4de8fe15d585" ], "author": { "name": "Jan Kiszka", "email": "jan.kiszka@siemens.com", "time": "Mon Aug 25 18:07:10 2025 +0200" }, "committer": { "name": "Ard Biesheuvel", "email": "ardb@kernel.org", "time": "Tue Aug 26 14:47:53 2025 +0200" }, "message": "efi: stmm: Fix incorrect buffer allocation method\n\nThe communication buffer allocated by setup_mm_hdr() is later on passed\nto tee_shm_register_kernel_buf(). The latter expects those buffers to be\ncontiguous pages, but setup_mm_hdr() just uses kmalloc(). That can cause\nvarious corruptions or BUGs, specifically since commit 9aec2fb0fd5e\n(\"slab: allocate frozen pages\"), though it was broken before as well.\n\nFix this by using alloc_pages_exact() instead of kmalloc().\n\nFixes: c44b6be62e8d (\"efi: Add tee-based EFI variable driver\")\nSigned-off-by: Jan Kiszka \u003cjan.kiszka@siemens.com\u003e\nAcked-by: Ilias Apalodimas \u003cilias.apalodimas@linaro.org\u003e\nAcked-by: Sumit Garg \u003csumit.garg@oss.qualcomm.com\u003e\nSigned-off-by: Ard Biesheuvel \u003cardb@kernel.org\u003e\n" }, { "commit": "1f3214aae9f49faf495f3836216afbc6c5400b2e", "tree": "b98d3644113c9385050183ab8746be0dc8411337", "parents": [ "832e5777143e799a97e8f9b96f002a90f06ba548" ], "author": { "name": "Antheas Kapenekakis", "email": "lkml@antheas.dev", "time": "Sun Aug 03 18:02:53 2025 +0200" }, "committer": { "name": "Jiri Kosina", "email": "jkosina@suse.com", "time": "Tue Aug 26 12:48:58 2025 +0200" }, "message": "HID: quirks: add support for Legion Go dual dinput modes\n\nThe Legion Go features detachable controllers which support a dual\ndinput mode. In this mode, the controllers appear under a single HID\ndevice with two applications.\n\nCurrently, both controllers appear under the same event device, causing\ntheir controls to be mixed up. This patch separates the two so that\nthey can be used independently.\n\nIn addition, the latest firmware update for the Legion Go swaps the IDs\nto the ones used by the Legion Go 2, so add those IDs as well.\n\n[jkosina@suse.com: improved shortlog]\nSigned-off-by: Antheas Kapenekakis \u003clkml@antheas.dev\u003e\nSigned-off-by: Jiri Kosina \u003cjkosina@suse.com\u003e\n" } ], "next": "832e5777143e799a97e8f9b96f002a90f06ba548" }