Описание
In the Linux kernel, the following vulnerability has been resolved:
net: ethernet: cortina: Carry over frag counter
The gmac_rx() NAPI poll function assembles packets in an SKB from a ring buffer.
If the ring buffer gets completely emptied during a poll cycle, we exit gmac_rx(), but the packet is not yet completely assembled in the SKB, yet the fragment counter frag_nr is reset to zero on the next invocation.
Solve this by making the RX fragment counter a part of the port struct, and carry it over between invocations.
Reset the fragment counter only right after calling napi_gro_frags(), on error (after calling napi_free_frags()) or if stopping the port.
Reset it in some place where not strictly necessary just to emphasize what is going on.
This was found by Sashiko during normal patch review.
In the Linux kernel, the following vulnerability has been resolved:
net: ethernet: cortina: Carry over frag counter
The gmac_rx() NAPI poll function assembles packets in an SKB from a ring buffer.
If the ring buffer gets completely emptied during a poll cycle, we exit gmac_rx(), but the packet is not yet completely assembled in the SKB, yet the fragment counter frag_nr is reset to zero on the next invocation.
Solve this by making the RX fragment counter a part of the port struct, and carry it over between invocations.
Reset the fragment counter only right after calling napi_gro_frags(), on error (after calling napi_free_frags()) or if stopping the port.
Reset it in some place where not strictly necessary just to emphasize what is going on.
This was found by Sashiko during normal patch review.
Ссылки
- https://nvd.nist.gov/vuln/detail/CVE-2026-64055
- https://git.kernel.org/stable/c/46806096f35b8d3dfa2f321ddd77f597edcdb85f
- https://git.kernel.org/stable/c/7123cf481e21b54eb6adc4cb0d8dc2876aeaee41
- https://git.kernel.org/stable/c/75105fcf73f1ce7d9f769aaefec6e6d6645d5ac0
- https://git.kernel.org/stable/c/78cf08b3be47c28f07008a76c932bad7cdffa9d8
- https://git.kernel.org/stable/c/7af1fabdee744b7995fe01b30b77dfc397657cb5
- https://git.kernel.org/stable/c/c373b34877afea61c89e0dd2e38948c624249b9b
- https://git.kernel.org/stable/c/df31e3b64455293df1ea89c7da7d5c9bfbcdd253
- https://git.kernel.org/stable/c/ebd8ec2b309e3a447851b456ccaf8fb39f3661e7
Связанные уязвимости
In the Linux kernel, the following vulnerability has been resolved: net: ethernet: cortina: Carry over frag counter The gmac_rx() NAPI poll function assembles packets in an SKB from a ring buffer. If the ring buffer gets completely emptied during a poll cycle, we exit gmac_rx(), but the packet is not yet completely assembled in the SKB, yet the fragment counter frag_nr is reset to zero on the next invocation. Solve this by making the RX fragment counter a part of the port struct, and carry it over between invocations. Reset the fragment counter only right after calling napi_gro_frags(), on error (after calling napi_free_frags()) or if stopping the port. Reset it in some place where not strictly necessary just to emphasize what is going on. This was found by Sashiko during normal patch review.
In the Linux kernel, the following vulnerability has been resolved: net: ethernet: cortina: Carry over frag counter The gmac_rx() NAPI poll function assembles packets in an SKB from a ring buffer. If the ring buffer gets completely emptied during a poll cycle, we exit gmac_rx(), but the packet is not yet completely assembled in the SKB, yet the fragment counter frag_nr is reset to zero on the next invocation. Solve this by making the RX fragment counter a part of the port struct, and carry it over between invocations. Reset the fragment counter only right after calling napi_gro_frags(), on error (after calling napi_free_frags()) or if stopping the port. Reset it in some place where not strictly necessary just to emphasize what is going on. This was found by Sashiko during normal patch review.
In the Linux kernel, the following vulnerability has been resolved: net: ethernet: cortina: Carry over frag counter The gmac_rx() NAPI poll function assembles packets in an SKB from a ring buffer. If the ring buffer gets completely emptied during a poll cycle, we exit gmac_rx(), but the packet is not yet completely assembled in the SKB, yet the fragment counter frag_nr is reset to zero on the next invocation. Solve this by making the RX fragment counter a part of the port struct, and carry it over between invocations. Reset the fragment counter only right after calling napi_gro_frags(), on error (after calling napi_free_frags()) or if stopping the port. Reset it in some place where not strictly necessary just to emphasize what is going on. This was found by Sashiko during normal patch review.
In the Linux kernel, the following vulnerability has been resolved: n ...