mirror of
https://github.com/fkie-cad/nvd-json-data-feeds.git
synced 2025-05-28 09:11:28 +00:00
37 lines
3.0 KiB
JSON
37 lines
3.0 KiB
JSON
{
|
|
"id": "CVE-2021-47412",
|
|
"sourceIdentifier": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
|
|
"published": "2024-05-21T15:15:26.700",
|
|
"lastModified": "2024-11-21T06:36:05.480",
|
|
"vulnStatus": "Awaiting Analysis",
|
|
"cveTags": [],
|
|
"descriptions": [
|
|
{
|
|
"lang": "en",
|
|
"value": "In the Linux kernel, the following vulnerability has been resolved:\n\nblock: don't call rq_qos_ops->done_bio if the bio isn't tracked\n\nrq_qos framework is only applied on request based driver, so:\n\n1) rq_qos_done_bio() needn't to be called for bio based driver\n\n2) rq_qos_done_bio() needn't to be called for bio which isn't tracked,\nsuch as bios ended from error handling code.\n\nEspecially in bio_endio():\n\n1) request queue is referred via bio->bi_bdev->bd_disk->queue, which\nmay be gone since request queue refcount may not be held in above two\ncases\n\n2) q->rq_qos may be freed in blk_cleanup_queue() when calling into\n__rq_qos_done_bio()\n\nFix the potential kernel panic by not calling rq_qos_ops->done_bio if\nthe bio isn't tracked. This way is safe because both ioc_rqos_done_bio()\nand blkcg_iolatency_done_bio() are nop if the bio isn't tracked."
|
|
},
|
|
{
|
|
"lang": "es",
|
|
"value": "En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: bloque: no llame a rq_qos_ops->done_bio si no se realiza un seguimiento de la biograf\u00eda. El framework rq_qos solo se aplica en el controlador basado en solicitudes, por lo que: 1) rq_qos_done_bio() no necesita hacerlo ser llamado para un controlador basado en biograf\u00eda 2) No es necesario llamar a rq_qos_done_bio() para una biograf\u00eda que no est\u00e1 rastreada, como una biograf\u00eda terminada por un c\u00f3digo de manejo de errores. Especialmente en bio_endio(): 1) la cola de solicitudes se remite a trav\u00e9s de bio->bi_bdev->bd_disk->queue, que puede desaparecer ya que el recuento de la cola de solicitudes no se puede mantener en los dos casos anteriores 2) q->rq_qos se puede liberar en blk_cleanup_queue() al llamar a __rq_qos_done_bio() Solucione el posible p\u00e1nico del kernel al no llamar a rq_qos_ops->done_bio si no se realiza un seguimiento de la biograf\u00eda. Esta forma es segura porque tanto ioc_rqos_done_bio() como blkcg_iolatency_done_bio() no funcionan si no se realiza un seguimiento de la biograf\u00eda."
|
|
}
|
|
],
|
|
"metrics": {},
|
|
"references": [
|
|
{
|
|
"url": "https://git.kernel.org/stable/c/004b8f8a691205a93d9e80d98b786b2b97424d6e",
|
|
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"
|
|
},
|
|
{
|
|
"url": "https://git.kernel.org/stable/c/a647a524a46736786c95cdb553a070322ca096e3",
|
|
"source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67"
|
|
},
|
|
{
|
|
"url": "https://git.kernel.org/stable/c/004b8f8a691205a93d9e80d98b786b2b97424d6e",
|
|
"source": "af854a3a-2127-422b-91ae-364da2661108"
|
|
},
|
|
{
|
|
"url": "https://git.kernel.org/stable/c/a647a524a46736786c95cdb553a070322ca096e3",
|
|
"source": "af854a3a-2127-422b-91ae-364da2661108"
|
|
}
|
|
]
|
|
} |