FKIE_CVE-2025-71183

Vulnerability from fkie_nvd - Published: 2026-01-31 12:16 - Updated: 2026-02-03 16:44
Severity ?
Summary
In the Linux kernel, the following vulnerability has been resolved: btrfs: always detect conflicting inodes when logging inode refs After rename exchanging (either with the rename exchange operation or regular renames in multiple non-atomic steps) two inodes and at least one of them is a directory, we can end up with a log tree that contains only of the inodes and after a power failure that can result in an attempt to delete the other inode when it should not because it was not deleted before the power failure. In some case that delete attempt fails when the target inode is a directory that contains a subvolume inside it, since the log replay code is not prepared to deal with directory entries that point to root items (only inode items). 1) We have directories "dir1" (inode A) and "dir2" (inode B) under the same parent directory; 2) We have a file (inode C) under directory "dir1" (inode A); 3) We have a subvolume inside directory "dir2" (inode B); 4) All these inodes were persisted in a past transaction and we are currently at transaction N; 5) We rename the file (inode C), so at btrfs_log_new_name() we update inode C's last_unlink_trans to N; 6) We get a rename exchange for "dir1" (inode A) and "dir2" (inode B), so after the exchange "dir1" is inode B and "dir2" is inode A. During the rename exchange we call btrfs_log_new_name() for inodes A and B, but because they are directories, we don't update their last_unlink_trans to N; 7) An fsync against the file (inode C) is done, and because its inode has a last_unlink_trans with a value of N we log its parent directory (inode A) (through btrfs_log_all_parents(), called from btrfs_log_inode_parent()). 8) So we end up with inode B not logged, which now has the old name of inode A. At copy_inode_items_to_log(), when logging inode A, we did not check if we had any conflicting inode to log because inode A has a generation lower than the current transaction (created in a past transaction); 9) After a power failure, when replaying the log tree, since we find that inode A has a new name that conflicts with the name of inode B in the fs tree, we attempt to delete inode B... this is wrong since that directory was never deleted before the power failure, and because there is a subvolume inside that directory, attempting to delete it will fail since replay_dir_deletes() and btrfs_unlink_inode() are not prepared to deal with dir items that point to roots instead of inodes. When that happens the mount fails and we get a stack trace like the following: [87.2314] BTRFS info (device dm-0): start tree-log replay [87.2318] BTRFS critical (device dm-0): failed to delete reference to subvol, root 5 inode 256 parent 259 [87.2332] ------------[ cut here ]------------ [87.2338] BTRFS: Transaction aborted (error -2) [87.2346] WARNING: CPU: 1 PID: 638968 at fs/btrfs/inode.c:4345 __btrfs_unlink_inode+0x416/0x440 [btrfs] [87.2368] Modules linked in: btrfs loop dm_thin_pool (...) [87.2470] CPU: 1 UID: 0 PID: 638968 Comm: mount Tainted: G W 6.18.0-rc7-btrfs-next-218+ #2 PREEMPT(full) [87.2489] Tainted: [W]=WARN [87.2494] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.2-0-gea1b7a073390-prebuilt.qemu.org 04/01/2014 [87.2514] RIP: 0010:__btrfs_unlink_inode+0x416/0x440 [btrfs] [87.2538] Code: c0 89 04 24 (...) [87.2568] RSP: 0018:ffffc0e741f4b9b8 EFLAGS: 00010286 [87.2574] RAX: 0000000000000000 RBX: ffff9d3ec8a6cf60 RCX: 0000000000000000 [87.2582] RDX: 0000000000000002 RSI: ffffffff84ab45a1 RDI: 00000000ffffffff [87.2591] RBP: ffff9d3ec8a6ef20 R08: 0000000000000000 R09: ffffc0e741f4b840 [87.2599] R10: ffff9d45dc1fffa8 R11: 0000000000000003 R12: ffff9d3ee26d77e0 [87.2608] R13: ffffc0e741f4ba98 R14: ffff9d4458040800 R15: ffff9d44b6b7ca10 [87.2618] FS: 00007f7b9603a840(0000) GS:ffff9d4658982000(0000) knlGS:0000000000000000 [87. ---truncated---
References
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/0c2413c69129f6ce60157f7b53d9ba880260400b
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/7ba0b6461bc4edb3005ea6e00cdae189bcf908a5
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/a63998cd6687c14b160dccb0bbcf281b2eb0dab3
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/c7f0207db68d5a1b4af23acbef1a8e8ddc431ebb
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/d52af58dd463821c5c516aebb031a58934f696ea
Impacted products
Vendor Product Version
To clipboard 

{
  "cveTags": [],
  "descriptions": [
    {
      "lang": "en",
      "value": "In the Linux kernel, the following vulnerability has been resolved:\n\nbtrfs: always detect conflicting inodes when logging inode refs\n\nAfter rename exchanging (either with the rename exchange operation or\nregular renames in multiple non-atomic steps) two inodes and at least\none of them is a directory, we can end up with a log tree that contains\nonly of the inodes and after a power failure that can result in an attempt\nto delete the other inode when it should not because it was not deleted\nbefore the power failure. In some case that delete attempt fails when\nthe target inode is a directory that contains a subvolume inside it, since\nthe log replay code is not prepared to deal with directory entries that\npoint to root items (only inode items).\n\n1) We have directories \"dir1\" (inode A) and \"dir2\" (inode B) under the\n   same parent directory;\n\n2) We have a file (inode C) under directory \"dir1\" (inode A);\n\n3) We have a subvolume inside directory \"dir2\" (inode B);\n\n4) All these inodes were persisted in a past transaction and we are\n   currently at transaction N;\n\n5) We rename the file (inode C), so at btrfs_log_new_name() we update\n   inode C\u0027s last_unlink_trans to N;\n\n6) We get a rename exchange for \"dir1\" (inode A) and \"dir2\" (inode B),\n   so after the exchange \"dir1\" is inode B and \"dir2\" is inode A.\n   During the rename exchange we call btrfs_log_new_name() for inodes\n   A and B, but because they are directories, we don\u0027t update their\n   last_unlink_trans to N;\n\n7) An fsync against the file (inode C) is done, and because its inode\n   has a last_unlink_trans with a value of N we log its parent directory\n   (inode A) (through btrfs_log_all_parents(), called from\n   btrfs_log_inode_parent()).\n\n8) So we end up with inode B not logged, which now has the old name\n   of inode A. At copy_inode_items_to_log(), when logging inode A, we\n   did not check if we had any conflicting inode to log because inode\n   A has a generation lower than the current transaction (created in\n   a past transaction);\n\n9) After a power failure, when replaying the log tree, since we find that\n   inode A has a new name that conflicts with the name of inode B in the\n   fs tree, we attempt to delete inode B... this is wrong since that\n   directory was never deleted before the power failure, and because there\n   is a subvolume inside that directory, attempting to delete it will fail\n   since replay_dir_deletes() and btrfs_unlink_inode() are not prepared\n   to deal with dir items that point to roots instead of inodes.\n\n   When that happens the mount fails and we get a stack trace like the\n   following:\n\n   [87.2314] BTRFS info (device dm-0): start tree-log replay\n   [87.2318] BTRFS critical (device dm-0): failed to delete reference to subvol, root 5 inode 256 parent 259\n   [87.2332] ------------[ cut here ]------------\n   [87.2338] BTRFS: Transaction aborted (error -2)\n   [87.2346] WARNING: CPU: 1 PID: 638968 at fs/btrfs/inode.c:4345 __btrfs_unlink_inode+0x416/0x440 [btrfs]\n   [87.2368] Modules linked in: btrfs loop dm_thin_pool (...)\n   [87.2470] CPU: 1 UID: 0 PID: 638968 Comm: mount Tainted: G        W           6.18.0-rc7-btrfs-next-218+ #2 PREEMPT(full)\n   [87.2489] Tainted: [W]=WARN\n   [87.2494] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.2-0-gea1b7a073390-prebuilt.qemu.org 04/01/2014\n   [87.2514] RIP: 0010:__btrfs_unlink_inode+0x416/0x440 [btrfs]\n   [87.2538] Code: c0 89 04 24 (...)\n   [87.2568] RSP: 0018:ffffc0e741f4b9b8 EFLAGS: 00010286\n   [87.2574] RAX: 0000000000000000 RBX: ffff9d3ec8a6cf60 RCX: 0000000000000000\n   [87.2582] RDX: 0000000000000002 RSI: ffffffff84ab45a1 RDI: 00000000ffffffff\n   [87.2591] RBP: ffff9d3ec8a6ef20 R08: 0000000000000000 R09: ffffc0e741f4b840\n   [87.2599] R10: ffff9d45dc1fffa8 R11: 0000000000000003 R12: ffff9d3ee26d77e0\n   [87.2608] R13: ffffc0e741f4ba98 R14: ffff9d4458040800 R15: ffff9d44b6b7ca10\n   [87.2618] FS:  00007f7b9603a840(0000) GS:ffff9d4658982000(0000) knlGS:0000000000000000\n   [87.\n---truncated---"
    },
    {
      "lang": "es",
      "value": "En el kernel de Linux, la siguiente vulnerabilidad ha sido resuelta:  btrfs: siempre detectar inodos en conflicto al registrar referencias de inodos  Despu\u00e9s de un intercambio de renombre (ya sea con la operaci\u00f3n de intercambio de renombre o renombres regulares en m\u00faltiples pasos no at\u00f3micos) de dos inodos y al menos uno de ellos es un directorio, podemos terminar con un \u00e1rbol de registro que contiene solo uno de los inodos y despu\u00e9s de un fallo de energ\u00eda eso puede resultar en un intento de eliminar el otro inodo cuando no deber\u00eda porque no fue eliminado antes del fallo de energ\u00eda. En algunos casos, ese intento de eliminaci\u00f3n falla cuando el inodo de destino es un directorio que contiene un subvolumen dentro de \u00e9l, ya que el c\u00f3digo de reproducci\u00f3n del registro no est\u00e1 preparado para manejar entradas de directorio que apuntan a elementos ra\u00edz (solo elementos de inodo).  1) Tenemos directorios \u0027dir1\u0027 (inodo A) y \u0027dir2\u0027 (inodo B) bajo el mismo directorio padre;  2) Tenemos un archivo (inodo C) bajo el directorio \u0027dir1\u0027 (inodo A);  3) Tenemos un subvolumen dentro del directorio \u0027dir2\u0027 (inodo B);  4) Todos estos inodos fueron persistidos en una transacci\u00f3n pasada y actualmente estamos en la transacci\u00f3n N;  5) Renombramos el archivo (inodo C), as\u00ed que en btrfs_log_new_name() actualizamos el last_unlink_trans del inodo C a N;  6) Obtenemos un intercambio de renombre para \u0027dir1\u0027 (inodo A) y \u0027dir2\u0027 (inodo B), as\u00ed que despu\u00e9s del intercambio \u0027dir1\u0027 es el inodo B y \u0027dir2\u0027 es el inodo A. Durante el intercambio de renombre llamamos a btrfs_log_new_name() para los inodos A y B, pero como son directorios, no actualizamos su last_unlink_trans a N;  7) Se realiza un fsync contra el archivo (inodo C), y debido a que su inodo tiene un last_unlink_trans con un valor de N registramos su directorio padre (inodo A) (a trav\u00e9s de btrfs_log_all_parents(), llamado desde btrfs_log_inode_parent()).  8) As\u00ed que terminamos con el inodo B no registrado, que ahora tiene el nombre antiguo del inodo A. En copy_inode_items_to_log(), al registrar el inodo A, no verificamos si ten\u00edamos alg\u00fan inodo en conflicto para registrar porque el inodo A tiene una generaci\u00f3n inferior a la transacci\u00f3n actual (creado en una transacci\u00f3n pasada);  9) Despu\u00e9s de un fallo de energ\u00eda, al reproducir el \u00e1rbol de registro, ya que encontramos que el inodo A tiene un nuevo nombre que entra en conflicto con el nombre del inodo B en el \u00e1rbol del sistema de archivos, intentamos eliminar el inodo B... esto es incorrecto ya que ese directorio nunca fue eliminado antes del fallo de energ\u00eda, y porque hay un subvolumen dentro de ese directorio, intentar eliminarlo fallar\u00e1 ya que replay_dir_deletes() y btrfs_unlink_inode() no est\u00e1n preparados para manejar elementos de directorio que apuntan a ra\u00edces en lugar de inodos.  Cuando eso sucede el montaje falla y obtenemos un seguimiento de pila como el siguiente:  [87.2314] Informaci\u00f3n de BTRFS (dispositivo dm-0): inicio de reproducci\u00f3n del registro de \u00e1rbol [87.2318] Cr\u00edtico de BTRFS (dispositivo dm-0): fall\u00f3 al eliminar la referencia al subvolumen, ra\u00edz 5 inodo 256 padre 259 [87.2332] ------------[ cut here ]------------ [87.2338] BTRFS: Transacci\u00f3n abortada (error -2) [87.2346] ADVERTENCIA: CPU: 1 PID: 638968 en fs/btrfs/inode.c:4345 __btrfs_unlink_inode+0x416/0x440 [btrfs] [87.2368] M\u00f3dulos enlazados: btrfs loop dm_thin_pool (...) [87.2470] CPU: 1 UID: 0 PID: 638968 Comm: mount Tainted: G        W           6.18.0-rc7-btrfs-next-218+ #2 PREEMPT(full) [87.2489] Tainted: [W]=WARN [87.2494] Nombre del hardware: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.16.2-0-gea1b7a073390-prebuilt.qemu.org 04/01/2014 [87.2514] RIP: 0010:__btrfs_unlink_inode+0x416/0x440 [btrfs] [87.2538] C\u00f3digo: c0 89 04 24 (...) [87.2568] RSP: 0018:ffffc0e741f4b9b8 EFLAGS: 00010286 [87.2574] RAX: 0000000000000000 RBX: ffff9d3ec8a6cf60 RCX: 0000000000000000 [87.2582] RDX: 0000000000000002 RSI: ffffffff84ab45a1 RDI: 00000000ffffffff [87.2591] RBP: ffff9d3ec8a6ef20 R08: "
    }
  ],
  "id": "CVE-2025-71183",
  "lastModified": "2026-02-03T16:44:36.630",
  "metrics": {},
  "published": "2026-01-31T12:16:03.570",
  "references": [
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/0c2413c69129f6ce60157f7b53d9ba880260400b"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/7ba0b6461bc4edb3005ea6e00cdae189bcf908a5"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/a63998cd6687c14b160dccb0bbcf281b2eb0dab3"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/c7f0207db68d5a1b4af23acbef1a8e8ddc431ebb"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/d52af58dd463821c5c516aebb031a58934f696ea"
    }
  ],
  "sourceIdentifier": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
  "vulnStatus": "Awaiting Analysis"
}


Log in or create an account to share your comment.




Tags
Taxonomy of the tags.




Sightings

Author Source Type Date

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.



Detection rules are retrieved from Rulezet.