** Tags added: kernel-daily-bug -- You received this bug notification because you are subscribed to linux in Ubuntu. Matching subscriptions: Bgg, Bmail, Nb https://bugs.launchpad.net/bugs/2165844 Title: linux 7.0.0-30: kernel BUG at fs/iomap/buffered-io.c:1061 in iomap_write_end() on ntfs3 buffered write Status in linux package in Ubuntu: New Bug description: [Summary] Any buffered write() to an ntfs3-mounted volume can hit a BUG_ON in the generic iomap write path and panic the kernel. An unprivileged process (rsync, uid 1000) took the whole machine down; the system froze instantly, then kdump rebooted into the capture kernel. [Impact] This is reachable from unprivileged userspace with a plain write(2) syscall. Whatever the state of the NTFS volume, the correct outcome is -EIO returned to the caller, not a kernel panic. Reproducing it needs nothing but a writable NTFS mount, which udisks/GNOME Files creates by default when a user clicks a Windows partition. [Environment] Ubuntu 26.04, linux-image-7.0.0-30-generic 7.0.0-30.30, x86_64 ASUS ROG STRIX B550-XE GAMING WIFI, BIOS 3607 NTFS volume: 1.8 TB Windows system partition, mounted read-write by udisks (driver ntfs3, POSIX ACLs enabled) [What happened] Writing to the NTFS mount, the kernel hit the assertion and died: kernel BUG at fs/iomap/buffered-io.c:1061! Oops: invalid opcode: 0000 [#1] SMP NOPTI CPU: 2 UID: 1000 PID: 18561 Comm: rsync Kdump: loaded Tainted: G O Hardware name: ASUS System Product Name/ROG STRIX B550-XE GAMING WIFI RIP: 0010:iomap_write_end+0x1e0/0x1f0 Call Trace: iomap_write_iter+0x171/0x340 iomap_file_buffered_write+0xa6/0x110 ntfs_file_write_iter+0x267/0x310 [ntfs3] vfs_write+0x25b/0x490 ksys_write+0x71/0xf0 __x64_sys_write+0x19/0x30 do_syscall_64+0x105/0x5a0 entry_SYSCALL_64_after_hwframe+0x76/0x7e The faulting syscall was a 137-byte write(2) to fd 1 (ORIG_RAX 0x1, RDX 0x89) — an ordinary small buffered write, not anything exotic. The G/O taint is the out-of-tree NVIDIA module (595.84). It is unrelated to this code path; ntfs3, iomap and the write path are all in-tree. [Preceding symptom] About 14 minutes before the panic, ntfs3 logged a large burst of metadata repairs on the same volume: ntfs3(nvme0n1p2): ino=..., Correct links count -> N ntfs3: 15286 callbacks suppressed So the volume was not perfectly clean. That may well be what steers the write path into the bad state — but an inconsistent filesystem must not be able to panic the kernel, so it looks like a missing error path rather than a disk problem. [Reproducer] Not deterministic here, but the shape is simple: 1. Mount an NTFS volume read-write with ntfs3 (udisksctl mount is enough). 2. Run sustained buffered I/O against it (rsync of a large tree). The other reports linked below describe the same trace from ordinary file writes, including games writing save data. [Already reported elsewhere] Same trace, same line, on 7.0.x kernels since roughly April 2026: - https://github.com/CachyOS/linux-cachyos/issues/841 - https://discuss.cachyos.org/t/kernel-bug-in-iomap-write-end-triggered-by-ntfs3-buffered-write-linux-7-0-1-cachyos/28546 - https://bbs.archlinux.org/viewtopic.php?pid=2296728 - https://forum.manjaro.org/t/total-system-hangs-and-fs-iomap-kernel-panic-on-7-0-10-stable-on-6-18-lts/188283 I could not find an upstream fix. Upstream appears to be replacing ntfs3 with the new ntfsplus driver in 7.1 rather than fixing this one. [Suggestion for Ubuntu] Since 26.04 ships 7.0 and a user can trigger a kernel panic by copying files onto a Windows partition from the file manager, it may be worth either backporting a fix or having udisks default NTFS mounts to read-only until 7.1 lands via HWE. [Attached] Full kernel ring buffer from the kdump vmcore (VmCoreDmesg), plus the usual apport-collected hardware and package data. Sanitisation applied before upload: firewall log lines were dropped, and MAC addresses, the wireless BSSID/SSID and LAN IPv4 addresses were replaced with placeholders in VmCoreDmesg, CurrentDmesg and WifiSyslog. Nothing else was altered; the panic trace is untouched. An incomplete vmcore (2.3 GB, truncated because the machine was power-cycled during the dump) is available if it would help. ProblemType: Bug DistroRelease: Ubuntu 26.04 Package: linux-image-7.0.0-30-generic 7.0.0-30.30 ProcVersionSignature: Ubuntu 7.0.0-30.30-generic 7.0.12 Uname: Linux 7.0.0-30-generic x86_64 ApportVersion: 2.34.1-0ubuntu0.1 Architecture: amd64 CasperMD5CheckResult: pass Date: Sun Aug 30 11:45:03 2026 InstallationDate: Installed on 2026-07-12 (49 days ago) InstallationMedia: Ubuntu 26.04 "Resolute Raccoon" - Release amd64 (20260423.1) MachineType: ASUS System Product Name PackageArchitecture: amd64 ProcFB: 0 nvidia-drmdrmfb ProcKernelCmdLine: BOOT_IMAGE=/boot/vmlinuz-7.0.0-30-generic root=UUID=5778b2dd-b496-4525-b52f-223e5be557bc ro quiet splash crashkernel=2G-4G:320M,4G-32G:512M,32G-64G:1024M,64G-128G:2048M,128G-:4096M SourcePackage: linux Title: linux 7.0.0-30: kernel BUG at fs/iomap/buffered-io.c:1061 in iomap_write_end() on ntfs3 buffered write UpgradeStatus: No upgrade log present (probably fresh install) _MarkForUpload: True dmi.bios.date: 03/18/2024 dmi.bios.release: 5.17 dmi.bios.vendor: American Megatrends Inc. dmi.bios.version: 3607 dmi.board.asset.tag: Default string dmi.board.name: ROG STRIX B550-XE GAMING WIFI dmi.board.vendor: ASUSTeK COMPUTER INC. dmi.board.version: Rev X.0x dmi.chassis.asset.tag: Default string dmi.chassis.type: 3 dmi.chassis.vendor: Default string dmi.chassis.version: Default string dmi.modalias: dmi:bvnAmericanMegatrendsInc.:bvr3607:bd03/18/2024:br5.17:svnASUS:pnSystemProductName:pvrSystemVersion:rvnASUSTeKCOMPUTERINC.:rnROGSTRIXB550-XEGAMINGWIFI:rvrRevX.0x:cvnDefaultstring:ct3:cvrDefaultstring:skuSKU:pfaTobefilledbyO.E.M.: dmi.product.family: To be filled by O.E.M. dmi.product.name: System Product Name dmi.product.sku: SKU dmi.product.version: System Version dmi.sys.vendor: ASUS To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2165844/+subscriptions
Комментариев нет:
Отправить комментарий