fkie_cve-2022-49767
Vulnerability from fkie_nvd
Published
2025-05-01 15:15
Modified
2025-05-02 13:53
Severity ?
Summary
In the Linux kernel, the following vulnerability has been resolved: 9p/trans_fd: always use O_NONBLOCK read/write syzbot is reporting hung task at p9_fd_close() [1], for p9_mux_poll_stop() from p9_conn_destroy() from p9_fd_close() is failing to interrupt already started kernel_read() from p9_fd_read() from p9_read_work() and/or kernel_write() from p9_fd_write() from p9_write_work() requests. Since p9_socket_open() sets O_NONBLOCK flag, p9_mux_poll_stop() does not need to interrupt kernel_read()/kernel_write(). However, since p9_fd_open() does not set O_NONBLOCK flag, but pipe blocks unless signal is pending, p9_mux_poll_stop() needs to interrupt kernel_read()/kernel_write() when the file descriptor refers to a pipe. In other words, pipe file descriptor needs to be handled as if socket file descriptor. We somehow need to interrupt kernel_read()/kernel_write() on pipes. A minimal change, which this patch is doing, is to set O_NONBLOCK flag from p9_fd_open(), for O_NONBLOCK flag does not affect reading/writing of regular files. But this approach changes O_NONBLOCK flag on userspace- supplied file descriptors (which might break userspace programs), and O_NONBLOCK flag could be changed by userspace. It would be possible to set O_NONBLOCK flag every time p9_fd_read()/p9_fd_write() is invoked, but still remains small race window for clearing O_NONBLOCK flag. If we don't want to manipulate O_NONBLOCK flag, we might be able to surround kernel_read()/kernel_write() with set_thread_flag(TIF_SIGPENDING) and recalc_sigpending(). Since p9_read_work()/p9_write_work() works are processed by kernel threads which process global system_wq workqueue, signals could not be delivered from remote threads when p9_mux_poll_stop() from p9_conn_destroy() from p9_fd_close() is called. Therefore, calling set_thread_flag(TIF_SIGPENDING)/recalc_sigpending() every time would be needed if we count on signals for making kernel_read()/kernel_write() non-blocking. [Dominique: add comment at Christian's suggestion]
References
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/0b5e6bd72b8171364616841603a70e4ba9837063
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/0e07032b4b4724b8ad1003698cb81083c1818999
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/5af16182c5639349415118e9e9aecd8355f7a08b
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/7abf40f06a76c0dff42eada10597917e9776fbd4
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/9f8554615df668e4bf83294633ee9d232b28ce45
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/a8e2fc8f7b41fa9d9ca5f624f4e4d34fce5b40a9
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/b1ad04da7fe4515e2ce2d5f2dcab3b5b6d45614b
416baaa9-dc9f-4396-8d5f-8c081fb06d67https://git.kernel.org/stable/c/ef575281b21e9a34dfae544a187c6aac2ae424a9
Impacted products
Vendor Product Version


To clipboard 

{
  "cveTags": [],
  "descriptions": [
    {
      "lang": "en",
      "value": "In the Linux kernel, the following vulnerability has been resolved:\n\n9p/trans_fd: always use O_NONBLOCK read/write\n\nsyzbot is reporting hung task at p9_fd_close() [1], for p9_mux_poll_stop()\n from p9_conn_destroy() from p9_fd_close() is failing to interrupt already\nstarted kernel_read() from p9_fd_read() from p9_read_work() and/or\nkernel_write() from p9_fd_write() from p9_write_work() requests.\n\nSince p9_socket_open() sets O_NONBLOCK flag, p9_mux_poll_stop() does not\nneed to interrupt kernel_read()/kernel_write(). However, since p9_fd_open()\ndoes not set O_NONBLOCK flag, but pipe blocks unless signal is pending,\np9_mux_poll_stop() needs to interrupt kernel_read()/kernel_write() when\nthe file descriptor refers to a pipe. In other words, pipe file descriptor\nneeds to be handled as if socket file descriptor.\n\nWe somehow need to interrupt kernel_read()/kernel_write() on pipes.\n\nA minimal change, which this patch is doing, is to set O_NONBLOCK flag\n from p9_fd_open(), for O_NONBLOCK flag does not affect reading/writing\nof regular files. But this approach changes O_NONBLOCK flag on userspace-\nsupplied file descriptors (which might break userspace programs), and\nO_NONBLOCK flag could be changed by userspace. It would be possible to set\nO_NONBLOCK flag every time p9_fd_read()/p9_fd_write() is invoked, but still\nremains small race window for clearing O_NONBLOCK flag.\n\nIf we don\u0027t want to manipulate O_NONBLOCK flag, we might be able to\nsurround kernel_read()/kernel_write() with set_thread_flag(TIF_SIGPENDING)\nand recalc_sigpending(). Since p9_read_work()/p9_write_work() works are\nprocessed by kernel threads which process global system_wq workqueue,\nsignals could not be delivered from remote threads when p9_mux_poll_stop()\n from p9_conn_destroy() from p9_fd_close() is called. Therefore, calling\nset_thread_flag(TIF_SIGPENDING)/recalc_sigpending() every time would be\nneeded if we count on signals for making kernel_read()/kernel_write()\nnon-blocking.\n\n[Dominique: add comment at Christian\u0027s suggestion]"
    },
    {
      "lang": "es",
      "value": "En el kernel de Linux, se ha resuelto la siguiente vulnerabilidad: 9p/trans_fd: syzbot reporta que la tarea se bloquea en p9_fd_close() [1], ya que p9_mux_poll_stop() de p9_conn_destroy() de p9_fd_close() no interrumpe las solicitudes kernel_read() de p9_fd_read() de p9_read_work() o kernel_write() de p9_fd_write() de p9_write_work() ya iniciadas. Dado que p9_socket_open() establece el indicador O_NONBLOCK, p9_mux_poll_stop() no necesita interrumpir kernel_read()/kernel_write(). Sin embargo, dado que p9_fd_open() no establece el indicador O_NONBLOCK, sino que bloquea la tuber\u00eda a menos que la se\u00f1al est\u00e9 pendiente, p9_mux_poll_stop() necesita interrumpir kernel_read()/kernel_write() cuando el descriptor de archivo hace referencia a una tuber\u00eda. En otras palabras, el descriptor de archivo de la tuber\u00eda debe manejarse como si fuera un descriptor de archivo de socket. De alguna manera, necesitamos interrumpir kernel_read()/kernel_write() en las tuber\u00edas. Un cambio m\u00ednimo, que este parche est\u00e1 realizando, es establecer el indicador O_NONBLOCK de p9_fd_open(), ya que el indicador O_NONBLOCK no afecta la lectura/escritura de archivos normales. Pero este enfoque cambia el indicador O_NONBLOCK en los descriptores de archivo proporcionados por el espacio de usuario (lo que podr\u00eda romper los programas del espacio de usuario), y el indicador O_NONBLOCK podr\u00eda ser cambiado por el espacio de usuario. Ser\u00eda posible establecer el indicador O_NONBLOCK cada vez que se invoca p9_fd_read()/p9_fd_write(), pero a\u00fan queda una peque\u00f1a ventana de tiempo para borrar el indicador O_NONBLOCK. Si no queremos manipular el indicador O_NONBLOCK, podr\u00edamos rodear kernel_read()/kernel_write() con set_thread_flag(TIF_SIGPENDING) y recalc_sigpending(). Dado que los trabajos de p9_read_work()/p9_write_work() son procesados por hilos del n\u00facleo que procesan la cola de trabajo global system_wq, no se pudieron enviar se\u00f1ales desde hilos remotos al llamar a p9_mux_poll_stop() de p9_conn_destroy() de p9_fd_close(). Por lo tanto, ser\u00eda necesario llamar a set_thread_flag(TIF_SIGPENDING)/recalc_sigpending() cada vez si dependemos de se\u00f1ales para que kernel_read()/kernel_write() sea no bloqueante. [Dominique: a\u00f1adir comentario a la sugerencia de Christian]"
    }
  ],
  "id": "CVE-2022-49767",
  "lastModified": "2025-05-02T13:53:20.943",
  "metrics": {},
  "published": "2025-05-01T15:15:59.483",
  "references": [
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/0b5e6bd72b8171364616841603a70e4ba9837063"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/0e07032b4b4724b8ad1003698cb81083c1818999"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/5af16182c5639349415118e9e9aecd8355f7a08b"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/7abf40f06a76c0dff42eada10597917e9776fbd4"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/9f8554615df668e4bf83294633ee9d232b28ce45"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/a8e2fc8f7b41fa9d9ca5f624f4e4d34fce5b40a9"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/b1ad04da7fe4515e2ce2d5f2dcab3b5b6d45614b"
    },
    {
      "source": "416baaa9-dc9f-4396-8d5f-8c081fb06d67",
      "url": "https://git.kernel.org/stable/c/ef575281b21e9a34dfae544a187c6aac2ae424a9"
    }
  ],
  "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 seen somewhere by the user.
  • Confirmed: The vulnerability is confirmed from an analyst perspective.
  • Exploited: This vulnerability was exploited and seen by the user reporting the sighting.
  • Patched: This vulnerability was successfully patched by the user reporting the sighting.
  • Not exploited: This vulnerability was not exploited or seen by the user reporting the sighting.
  • Not confirmed: The user expresses doubt about the veracity of the vulnerability.
  • Not patched: This vulnerability was not successfully patched by the user reporting the sighting.