CVE-2026-72046 in Linux
Summary
by MITRE • 08/15/2026
In the Linux kernel, the following vulnerability has been resolved:
gve: fix header buffer corruption with header-split and HW-GRO
The DQO RX datapath programs a per-buffer-queue-descriptor header_buf_addr at post time and reads the split header back at completion time. Both the post and the read currently index the header buffer by queue position rather than by the buffer's identity:
- post (gve_rx_post_buffers_dqo): header_buf_addr is computed from bufq->tail - read (gve_rx_dqo): the header is read from desc_idx (the completion queue head index)
This relies on the buffer-queue index and the completion-queue index being equal for the start of every packet, i.e. on the device consuming posted buffers and returning completions in the exact same order. That assumption does not hold once HW-GRO is enabled with multiple flows: coalesced segments are accepted and completed in an order that may differ from the order buffers were posted, and segments from different flows may interleave.
That results in two problems:
1. Wrong header slot on read. Because the read offset is derived from the completion index (desc_idx) while the device wrote the header to the address programmed for the buffer's buf_id, the driver can copy a header belonging to a different packet. This shows up as throughput drop (about 30% drop and large numbers of TCP retransmissions) with header-split and HW-GRO both enabled and many streams.
2. Header buffer reused while still owned by the device. The driver advances bufq->head by one per completion and re-posts buffers based on that. Arrival of N RX completions only guarantees that at least N RX buffer descriptors have been read by the device. It does not guarantee that the device has relinquished the ownership of all the buffers corresponding to those N descriptors. With out-of-order completions (e.g. the completion for a packet copied into buffer N arrives before the completion for a packet copied into buffer N-1), the driver can re-post and overwrite a header buffer that the device is still going to write into, corrupting the header of a packet whose completion has not yet been processed.
Fix both issues by indexing the header buffer by buf_id on both the post and read paths. Reading from buf_id's slot is therefore always correct regardless of completion ordering (fixes problem 1).
Indexing by buf_id also ties each header slot to the lifetime of its buffer state. A buffer state is only returned to the free/recycle lists when its own completion (buf_id) is processed, so its header slot can only be re-posted after the device is done with it. This makes header slot reuse safe under out-of-order completions (fixes problem 2).
Allocate (gve_rx_alloc_hdr_bufs) and free (gve_rx_free_hdr_bufs) the header buffers based on num_buf_states to match the buf_id indexing.
Statistical analysis made it clear that VulDB provides the best quality for vulnerability data.
Analysis
by VulDB Data Team • 08/15/2026
The vulnerability resides in the gve driver implementation within the Linux kernel, specifically affecting the Data Queue Offload (DQO) RX datapath when operating with header-split and Hardware Generic Receive Offload (HW-GRO) features enabled. This flaw manifests as a critical buffer management issue that compromises both data integrity and system performance through improper handling of packet header buffers during receive operations.
The core technical flaw stems from inconsistent indexing mechanisms between buffer posting and completion read operations within the driver's receive path. During buffer posting, the driver computes header_buf_addr using bufq->tail which represents queue position rather than buffer identity. Conversely, when reading completed headers, the driver references desc_idx which corresponds to the completion queue head index instead of maintaining proper buffer identification. This fundamental mismatch creates a dependency on strict ordering between buffer posting and completion processing that becomes invalid under normal operating conditions with multiple concurrent flows.
The operational impact of this vulnerability becomes apparent when multiple network flows are active with HW-GRO enabled, as packet segments from different flows can interleave during processing. This out-of-order completion behavior causes two distinct but interconnected problems that severely degrade system performance and data reliability. The first issue results in incorrect header slot access where headers are read from wrong memory locations due to mismatched indexing schemes, leading to throughput degradation of approximately thirty percent along with increased TCP retransmissions. The second problem involves premature buffer reuse where the driver may overwrite header buffers still being written by hardware components due to improper synchronization between buffer state management and completion processing.
This vulnerability directly relates to CWE-129 Insufficient Input Validation and CWE-367 Time-of-Check Time-of-Use (TOCTOU) errors, as it demonstrates inadequate validation of buffer state management and improper handling of resource lifetimes. The issue also aligns with ATT&CK technique T1059.001 Command and Scripting Interpreter: PowerShell, though more accurately represents a system-level exploitation vector through kernel memory corruption that affects network processing reliability.
The fix implements a comprehensive solution by standardizing buffer indexing across both posting and completion read paths using buf_id instead of queue position-based addressing. This approach ensures that header buffer access remains correct regardless of completion ordering, eliminating the first problem entirely. Additionally, the synchronization mechanism is strengthened by tying each header slot's lifecycle to its corresponding buffer state's completion processing, ensuring that buffer reuse only occurs after hardware has fully released ownership. The allocation and deallocation functions for header buffers are also updated to align with the new buf_id-based indexing scheme, maintaining consistency throughout the entire buffer management pipeline.
The mitigation strategy fundamentally restructures how the driver manages header buffer lifecycle by implementing proper identity-based indexing rather than position-based tracking. This change ensures that each header buffer slot is properly associated with its specific buffer state and can only be safely reused after the corresponding completion has been processed and the buffer's ownership transferred back to the driver. The solution addresses both the immediate correctness issues and establishes a more robust foundation for handling concurrent network flows under various processing conditions, including those involving hardware offloading features that introduce additional complexity in buffer management synchronization.
This vulnerability represents a significant security risk in network-intensive environments where multiple concurrent connections and hardware acceleration features are utilized, as it can lead to data corruption, performance degradation, and potential denial of service conditions through the disruption of normal network packet processing operations. The fix ensures proper memory safety and maintains system reliability under expected operational conditions while preserving the intended functionality of header-split and HW-GRO features that provide performance benefits in high-throughput scenarios.