wireguard-linux/include/linux, branch devel

ipv4: igmp: encode multicast exponential fields

2026-05-05T02:09:41Z

In IGMP, MRC and QQIC fields are not correctly encoded when generating query packets. Since the receiver of the query interprets these fields using the IGMPv3 floating- point decoding logic, any value that exceeds the linear threshold is incorrectly parsed as an exponential value, leading to an incorrect interval calculation. Encode and assign the corresponding protocol fields during query generation. Introduce the logic to dynamically calculate the exponent and mantissa using bit-scan (fls). This ensures MRC and QQIC fields (8-bit) are properly encoded when transmitting query packets with intervals that exceed their respective linear threshold value of 128 (for MRT/QQI). RFC3376: for both MRC and QQIC, values >= 128 represent the same floating-point encoding as follows: 0 1 2 3 4 5 6 7 +-+-+-+-+-+-+-+-+ |1| exp | mant | +-+-+-+-+-+-+-+-+ Reviewed-by: Nikolay Aleksandrov Reviewed-by: Ido Schimmel Signed-off-by: Ujjal Roy Link: https://patch.msgid.link/20260502131907.987-4-royujjal@gmail.com Signed-off-by: Jakub Kicinski

ipv4: igmp: get rid of IGMPV3_{QQIC,MRC} and simplify calculation

2026-05-05T02:09:41Z

Get rid of the IGMPV3_MRC macro and use the igmpv3_mrt() API to calculate the Max Resp Time from the Maximum Response Code. Similarly, for IGMPV3_QQIC, use the igmpv3_qqi() API to calculate the Querier's Query Interval from the QQIC field. Reviewed-by: Nikolay Aleksandrov Reviewed-by: Ido Schimmel Signed-off-by: Ujjal Roy Link: https://patch.msgid.link/20260502131907.987-2-royujjal@gmail.com Signed-off-by: Jakub Kicinski

ixgbe: E610: add discovering EEE capability

2026-05-03T02:12:36Z

Add detecting and parsing EEE device capability. Recently EEE functionality support has been introduced to E610 FW. Currently ixgbe driver has no possibility to detect whether NVM loaded on given adapter supports EEE. There's dedicated device capability element reflecting FW support for given EEE link speed. Reviewed-by: Aleksandr Loktionov Signed-off-by: Jedrzej Jagielski Tested-by: Rinitha S Signed-off-by: Jacob Keller Link: https://patch.msgid.link/20260430-jk-iwl-net-next-2026-04-30-v1-1-6f27ae1cd073@intel.com Signed-off-by: Jakub Kicinski

net/mlx5: use internal dma pools for frag buf alloc

2026-05-02T02:02:07Z

Add mlx5_dma_pool alloc/free paths, and wire mlx5_frag_buf allocation and free paths to use them. mlx5_frag_buf_alloc_node() now selects an mlx5_dma_pool to allocate fragments from, instead of directly allocating full coherent pages. mlx5_frag_buf_free() frees from the respective pool. mlx5_dma_pool_alloc() keeps allocation fast by maintaining pages with available indexes at the head of the list, so the common allocation path can take a free index immediately. New backing pages are allocated only when no free index is available. mlx5_dma_pool_free() returns released indexes to the pool and frees a backing page once all of its indexes become free. This avoids keeping fully free pages for the lifetime of the pool and reduces coherent DMA memory footprint. Signed-off-by: Nimrod Oren Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260429201429.223809-4-tariqt@nvidia.com Signed-off-by: Jakub Kicinski

net/mlx5: add frag buf pools create/destroy paths

2026-05-02T02:02:07Z

Introduce mlx5 DMA pool and pool-page data structures, and add the creation and teardown paths. Each NUMA node owns a set of mlx5_dma_pool instances, each one with a different block size. The sizes are defined as all powers of two starting from MLX5_ADAPTER_PAGE_SHIFT and up to PAGE_SHIFT. Since mlx5_frag_bufs are used to back objects whose sizes are encoded relative to MLX5_ADAPTER_PAGE_SHIFT, a smaller block_shift value cannot be used. Requests larger than PAGE_SIZE continue to be handled as page-sized fragments, as in the existing frag-buf allocation model. Signed-off-by: Nimrod Oren Signed-off-by: Tariq Toukan Link: https://patch.msgid.link/20260429201429.223809-3-tariqt@nvidia.com Signed-off-by: Jakub Kicinski

tcp: move max_packets_out, cwnd_usage_seq, rate_delivered and rate_interval_us to tcp_sock_write_tx group

2026-05-02T00:22:45Z

These fields are used in TX path. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260430100021.211139-6-edumazet@google.com Signed-off-by: Jakub Kicinski

tcp: move tp->bytes_acked to tcp_sock_write_tx group

2026-05-02T00:22:45Z

tp->bytes_acked is touched in TX path only. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260430100021.211139-5-edumazet@google.com Signed-off-by: Jakub Kicinski

tcp: move tp->first_tx_mstamp and tp->delivered_mstamp to tcp_sock_write_tx

2026-05-02T00:22:45Z

These fields are touched in when payload is sent. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260430100021.211139-4-edumazet@google.com Signed-off-by: Jakub Kicinski

tcp: move tp->segs_in and tp->segs_out to tcp_sock_write_txrx group

2026-05-02T00:22:45Z

segs_in is changed for each incoming packet, including ACK packets. segs_out is changed for each outgoing packet, including ACK packets. They belong to tcp_sock_write_txrx group. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260430100021.211139-3-edumazet@google.com Signed-off-by: Jakub Kicinski

tcp: move tp->delivered and tp->delivered_ce to tcp_sock_write_tx group

2026-05-02T00:22:44Z

These counters are changed whenever sent data is acknowleged. They do not belong to tcp_sock_write_txrx group, because TCP receivers do not touch them. Signed-off-by: Eric Dumazet Link: https://patch.msgid.link/20260430100021.211139-2-edumazet@google.com Signed-off-by: Jakub Kicinski