In the Linux kernel, the following vulnerability has been resolved: igb: remove napi_synchronize() in igb_down() When an AF_XDP zero-copy application terminates abruptly (e.g., kill -9), the XSK buffer pool is destroyed but NAPI polling continues. igb_clean_rx_irq_zc() repeatedly returns the full budget, preventing napi_complete_done() from clearing NAPI_STATE_SCHED. igb_down() calls napi_synchronize() before napi_disable() for each queue vector. napi_synchronize() spins waiting for NAPI_STATE_SCHED to clear, which never happens. igb_down() blocks indefinitely, the TX watchdog fires, and the TX queue remains permanently stalled. napi_disable() already handles this correctly: it sets NAPI_STATE_DISABLE. After a full-budget poll, __napi_poll() checks napi_disable_pending(). If set, it forces completion and clears NAPI_STATE_SCHED, breaking the loop that napi_synchronize() cannot. napi_synchronize() was added in commit 41f149a285da ("igb: Fix possible panic caused by Rx traffic arrival while interface is down"). napi_disable() provides stronger guarantees: it prevents further scheduling and waits for any active poll to exit. Other Intel drivers (ixgbe, ice, i40e) use napi_disable() without a preceding napi_synchronize() in their down paths. Remove redundant napi_synchronize() call and reorder napi_disable() before igb_set_queue_napi() so the queue-to-NAPI mapping is only cleared after polling has fully stopped.

A flaw was found in the Linux kernel's igb network driver. When an AF_XDP zero-copy application terminates abruptly, the network driver's NAPI polling mechanism can become stuck. This prevents the igb_down() function from completing, leading to the TX queue becoming permanently stalled. This can result in a Denial of Service (DoS) condition, making the network interface unresponsive.

igb can hang in igb_down when an AF_XDP zero copy application terminates abruptly and the XSK buffer pool is destroyed while NAPI polling keeps running. In this state the zero copy RX poll path can keep returning a full budget which prevents napi_complete_done from clearing NAPI_STATE_SCHED. igb_down then calls napi_synchronize and waits indefinitely for NAPI_STATE_SCHED to clear which never happens. This blocks the down path, triggers the TX watchdog, and can leave the TX queue permanently stalled which is a denial of service for the affected network device. The issue is not network reachable and requires local control of the host to set up and exercise AF_XDP zero copy. To mitigate this issue, prevent module igb from being loaded. Please see https://access.redhat.com/solutions/41278 for how to blacklist a kernel module to prevent it from loading automatically. Red Hat Enterprise Linux 10 Fix deferred kernel Red Hat Enterprise Linux 6 Not affected kernel Red Hat Enterprise Linux 7 Not affected kernel Red Hat Enterprise Linux 7 Not affected kernel-rt Red Hat Enterprise Linux 8 Not affected kernel Red Hat Enterprise Linux 8 Not affected kernel-rt Red Hat Enterprise Linux 9 Fix deferred kernel Red Hat Enterprise Linux 9 Fix deferred kernel-rt https://www.cve.org/CVERecord?id=CVE-2026-31691 https://nvd.nist.gov/vuln/detail/CVE-2026-31691 https://lore.kernel.org/linux-cve-announce/2026042757-CVE-2026-31691-5c48@gregkh/T