llamaCVEVulnerability Intelligence
/
Linux kernel

CVE-2026-99290

Linux kernel use-after-free in ksmbd after failed durable reconnect

An attacker could trigger writes to already-freed memory during session teardown, corrupting data.

Specific setups onlyParche disponibleksmbd
Descargar PNG JSON
CVSS 3.1
7.5
Alta
CVSS:3.1/AV:N/AC:H/PR:L/UI:N/S:U/C:H/I:H/A:H Fuente: CNA
ExplotaciónSin evidenciaNo figura en CISA KEV
Exploits públicos0Ninguno conocido
Parche1 versión6.7
Exposición4 dDesde la divulgación, 16 nov 2026, 4:59 a. m.
Detección8 minDesde la divulgación hasta su ingreso al índice
Timeline

Ciclo de vida de la vulnerabilidad

Hoy
Divulgación
Publicada
Modificada
Explotación
Remediación
Monitoreo
Detectada +2
Verificada
16 nov 26
19 nov 26
20 nov 26
16 nov 26
17 nov 26
18 nov 26
19 nov 26
20 nov 26
FechaFaseEventoDetalle
Divulgación Publicada CVE-2026-99290 divulgada · CVSS 7.5 (CNA)
Monitoreo Detectada Ingresó al índice
Monitoreo Fuente: NIST NVD Registro visto por primera vez en NIST NVD
Monitoreo Fuente: ENISA EUVD Registro visto por primera vez en ENISA EUVD
Divulgación Modificada Última actualización del registro
Monitoreo Verificada Última verificación contra las fuentes
Resumen
Resumen generado por IA · CC-BY 4.0 · 16 nov 2026

A use-after-free flaw in the ksmbd component of the Linux kernel occurs when a durable file handle reconnect fails. Tearing down the session then writes to memory that has already been freed. The description does not say how an attacker would trigger the failed reconnect.

Affects the Linux kernel's ksmbd SMB server from 6.6.32 up to 6.7, in the durable handle reconnect path. The fix is in 6.7.

¿Qué significa?A durable handle reconnect that fails can leave the handle published in the session file table. When the session is torn down, the kernel writes to freed memory, which can corrupt live kernel objects and crash the kernel.
¿Quién está afectado?The affected range is Linux kernel 6.6.32 up to but not including 6.7, and the bug needs ksmbd durable handle reconnects. Two overlapping reconnects of the same handle, where the later SMB2 CREATE fails after the handle is reopened, are the conditions that trigger it. Check the running kernel version with uname -r.
¿Qué hacer?Upgrade to Linux kernel 6.7 or later, since 6.7 is the listed fixed version. The facts give no vendor patch identifiers or workaround.
Descripción original del registro (Linux)

In the Linux kernel, the following vulnerability has been resolved: ksmbd: fix use-after-free after a failed durable reconnect ksmbd_reopen_durable_fd() republishes a detached durable handle in the reconnecting session's file table and takes no reference for that entry. If smb2_open() then fails, cleanup relies on the final ksmbd_fd_put() to remove the volatile id. Two reconnects of the same handle can both take a lookup reference before either attaches it. If the one that reopens first then fails and drops its references first, the other request makes the final put through ksmbd_put_durable_fd(). That frees the handle without removing the first session's table entry, so tearing that session down writes to freed memory. BUG: KASAN: slab-use-after-free in __close_file_table_ids+0x14d/0x740 [ksmbd] Call Trace: __close_file_table_ids+0x14d/0x740 [ksmbd] ksmbd_destroy_file_table+0x76/0x100 [ksmbd] ksmbd_session_destroy+0xf8/0x480 [ksmbd] ksmbd_dispose_sessions+0x13e/0x200 [ksmbd] ksmbd_conn_sessions_cleanup+0x4c3/0x6d0 [ksmbd] Freed by task 1030: __ksmbd_close_fd+0x97d/0x1000 [ksmbd] smb2_open+0x14eb/0xd9f0 [ksmbd] Commit 3515503322f4 ("ksmbd: fix durable reconnect error path file lifetime") handles a final put made by the reconnecting request, but not one made by another holder. Remove the volatile id under the file-table lock on this error path, before the references are dropped, as ksmbd_close_fd() does on the request_open path just above. Removal then no longer depends on which holder makes the final put. The stale comment in the reconnect cleanup is updated to match. A successful reconnect is unchanged.

Contexto

Condiciones y alcance

Condiciones necesarias para que el fallo sea explotable
01
A durable handle reconnect of a detached handle in ksmbd, where the reconnecting SMB2 CREATE republishes the handle into the session file table.“ksmbd_reopen_durable_fd() republishes a detached durable handle in the”
02
The subsequent smb2_open() call must fail after the handle is reopened.“If smb2_open() then fails, cleanup relies on the final ksmbd_fd_put() to”
03
Two reconnects of the same durable handle must overlap, with the second taking a lookup reference before the first attaches it.“Two reconnects of the same handle can both take a lookup reference before”
¿Quién debe revisar?Teams running Linux kernels with the ksmbd in-kernel SMB server and durable handles in use
Vector CVSS 3.1

Desglose de métricas

Opción resaltada = valor del vector · izquierda = peor caso · CNA
Vector de ataqueAV:N
RedAdyacenteLocalFísica
ComplejidadAC:H
BajaAlta
PrivilegiosPR:L
NingunoBajosAltos
InteracciónUI:N
NingunaRequerida
AlcanceS:U
CambiadoSin cambio
ConfidencialidadC:H
AltoBajoNinguno
IntegridadI:H
AltoBajoNinguno
DisponibilidadA:H
AltoBajoNinguno
Versiones

Rangos afectados

Afectada Corregida
Linux≥ 6.6.32 < 6.7
6.6.32
6.7
Remediación

Estado del parche por paquete

Paquete / productoEstadoCorregida enDesde divulgaciónÚltima versión
Linux
Corregida 6.7 — —
Versiones corregidas6.7
Explotación

Arsenal público

Sin exploits, módulos ni PoC públicos conocidos a la fecha de verificación.

Registro
EscenarioAV:N - A remote client sends SMB2 CREATE on the TCP connection accepted by ksmbd_tcp_new_connection and read in ksmbd_conn_handler_loop. parse_durable_handle_context() in smb2_open() takes the DH2C/DHnC persistent id from that request, and ksmbd_reopen_durable_fd() republishes the detached handle. AC:H - The stale volatile id is left behind only when smb2_open() fails after ksmbd_reopen_durable_fd() succeeds, in ksmbd_override_fsids() or ksmbd_vfs_getattr(), before request_open is set. The client can overlap a second lookup, but cannot force that credential allocation or getattr to fail. PR:L - __handle_ksmbd_work() requires smb2_check_user_session() and smb2_get_ksmbd_tcon() before smb2_open(), and the reopen that publishes the id must pass ksmbd_vfs_compare_durable_owner() in smb2_check_durable_oplock(). That is the authenticated handle owner, not a pre-auth CREATE. UI:N - The client opens its own durable handle, drops that connection so session_fd_check() preserves it, and sends the reconnecting SMB2 CREATEs. No other principal has to open a file or approve the reconnect. S:U - The dangling pointer is a ksmbd_file in the host ksmbd_file_cache, and ksmbd_session_destroy() corrupts that kernel object. The bug does not cross a guest/host or IOMMU boundary. C:H - ksmbd_put_durable_fd() frees the handle in __ksmbd_close_fd() with kmem_cache_free(). Session teardown then loads that idr pointer in __close_file_table_ids() and reads refcount, f_ci, and conn; a reused slot lets those reads follow attacker-influenced kernel pointers. I:H - __close_file_table_ids() writes the freed ksmbd_file, storing volatile_id, locking fp->f_ci, and running list_del_init() on fp->node. Once the ksmbd_file_cache slot is reused, that path corrupts a live object, including its connection reference. A:H - ksmbd_session_destroy() reaches the dangling id through ksmbd_destroy_file_table() and __close_file_table_ids() and touches the ksmbd_file already freed from smb2_open(), which is the slab-use-after-free that oopses the kernel.Linux
Referencias
Otrahttps://git.kernel.org/stable/c/5078a4e41b29acdf6be52e4a7cc8a59c2a72a471
Otrahttps://git.kernel.org/stable/c/bd47ee501b6ac4a26ae41f780a43b7967302a269
Otrahttps://git.kernel.org/stable/c/edacf65866a9f397438d5b9ff9f43312b29b4b3c
Otrahttps://git.kernel.org/stable/c/248b5819591c577727930e4e255a455235fab50a

Las referencias externas se muestran para copiar: cve.llama.ec no lo redirige a otros sitios.