Quick note before diving into the series details, Tariq is on vacation for a few days, I’ll be handling the mlx5 core/netdev submissions in his absence. This series adds support for zerocopy rx TCP with devmem and io_uring for ConnectX7 NICs and above. For performance reasons and simplicity HW-GRO will also be turned on when header-data split mode is on. Performance =========== Test setup: * CPU: Intel(R) Xeon(R) Platinum 8380 CPU @ 2.30GHz (single NUMA) * NIC: ConnectX7 * Benchmarking tool: kperf [0] * Single TCP flow * Test duration: 60s With application thread and interrupts pinned to the *same* core: |------+-----------+----------| | MTU | epoll | io_uring | |------+-----------+----------| | 1500 | 61.6 Gbps | 114 Gbps | | 4096 | 69.3 Gbps | 151 Gbps | | 9000 | 67.8 Gbps | 187 Gbps | |------+-----------+----------| The CPU usage for io_uring is 95%. Reproduction steps for io_uring: server --no-daemon -a 2001:db8::1 --no-memcmp --iou --iou_sendzc \ --iou_zcrx --iou_dev_name eth2 --iou_zcrx_queue_id 2 server --no-daemon -a 2001:db8::2 --no-memcmp --iou --iou_sendzc client --src 2001:db8::2 --dst 2001:db8::1 \ --msg-zerocopy -t 60 --cpu-min=2 --cpu-max=2 Patch overview: ================ First, a netmem API for skb_can_coalesce is added to the core to be able to do skb fragment coalescing on netmems. The next patches introduce some cleanups in the internal SHAMPO code and improvements to hw gro capability checks in FW. A separate page_pool is introduced for headers, to be used only when the rxq has a memory provider. Ethtool stats are added as well. Then the driver is converted to use the netmem API and to allow support for unreadable netmem page pool. The queue management ops are implemented. Finally, the tcp-data-split ring parameter is exposed. Changelog ========= Changes from v2 [2]: - Added support for netmem TX. - Changed skb_can_coalesce_netmem() based on Mina's suggestion. - Reworked tcp_data_split setting to no longer change HW-GRO in wanted_features. - Switched to a single page pool when rxq has no memory providers. Changes from v1 [1]: - Added support for skb_can_coalesce_netmem(). - Avoid netmem_to_page() casts in the driver. - Fixed code to abide 80 char limit with some exceptions to avoid code churn. References ========== [0] kperf: git://git.kernel.dk/kperf.git [1] v1: https://lore.kernel.org/all/20250116215530.158886-1-saeed@xxxxxxxxxx/ [2] v2: https://lore.kernel.org/all/1747950086-1246773-1-git-send-email-tariqt@xxxxxxxxxx/ Dragos Tatulea (3): net: Allow const args for of page_to_netmem() net: Add skb_can_coalesce for netmem net/mlx5e: Add TX support for netmems Saeed Mahameed (9): net/mlx5e: SHAMPO: Reorganize mlx5_rq_shampo_alloc net/mlx5e: SHAMPO: Remove redundant params net/mlx5e: SHAMPO: Improve hw gro capability checking net/mlx5e: SHAMPO: Separate pool for headers net/mlx5e: SHAMPO: Headers page pool stats net/mlx5e: Convert over to netmem net/mlx5e: Add support for UNREADABLE netmem page pools net/mlx5e: Implement queue mgmt ops and single channel swap net/mlx5e: Support ethtool tcp-data-split settings drivers/net/ethernet/mellanox/mlx5/core/en.h | 11 +- .../ethernet/mellanox/mlx5/core/en/params.c | 36 ++- .../net/ethernet/mellanox/mlx5/core/en/txrx.h | 3 +- .../ethernet/mellanox/mlx5/core/en_ethtool.c | 33 +- .../net/ethernet/mellanox/mlx5/core/en_main.c | 305 +++++++++++++----- .../net/ethernet/mellanox/mlx5/core/en_rx.c | 136 ++++---- .../ethernet/mellanox/mlx5/core/en_stats.c | 54 ++++ .../ethernet/mellanox/mlx5/core/en_stats.h | 24 ++ include/linux/skbuff.h | 12 +- include/net/netmem.h | 2 +- 10 files changed, 449 insertions(+), 167 deletions(-) base-commit: 2c7e4a2663a1ab5a740c59c31991579b6b865a26 -- 2.34.1
