Add support for providers with large rx buffer
Subsystems (15)
v9 cover Add support for providers with large rx buffer
Note: it's net/ only bits and doesn't include changes, which shoulf be merged separately and are posted separately. The full branch for convenience is at [1], and the patch is here:
Many modern NICs support configurable receive buffer lengths, and zcrx and memory providers can use buffers larger than 4K to improve performance. When paired with hw-gro larger rx buffer sizes can drastically reduce the number of buffers traversing the stack and save a lot of processing time. It also allows to give to users larger contiguous chunks of data. The idea was first floated around by Saeed during netdev conf 2024 and was asked about by a few folks.
Single stream benchmarks showed up to ~30% CPU util improvement. E.g. comparison for 4K vs 32K buffers using a 200Gbit NIC:
packets=23987040 (MB=2745098), rps=199559 (MB/s=22837)
CPU %usr %nice %sys %iowait %irq %soft %idle 0 1.53 0.00 27.78 2.72 1.31 66.45 0.22 packets=24078368 (MB=2755550), rps=200319 (MB/s=22924) CPU %usr %nice %sys %iowait %irq %soft %idle 0 0.69 0.00 8.26 31.65 1.83 57.00 0.57
This series adds net infrastructure for memory providers configuring the size and implements it for bnxt. It's an opt-in feature for drivers, they should advertise support for the parameter in the qops and must check if the hardware supports the given size. It's limited to memory providers as it drastically simplifies implementation. It doesn't affect the fast path zcrx uAPI, and the user exposed parameter is defined in zcrx terms, which allows it to be flexible and adjusted in the future.
A liburing example can be found at [2]
full branch:
[1] https://github.com/isilence/linux.git zcrx/large-buffers-v8 Liburing example:
[2] https://github.com/isilence/liburing.git zcrx/rx-buf-len
---
The following changes since commit 0f61b1860cc3f52aef9036d7235ed1f017632193:
Linux 6.19-rc5 (2026-01-11 17:03:14 -1000)
are available in the Git repository at:
https://github.com/isilence/linux.git tags/net-queue-rx-buf-len-v9
for you to fetch changes up to d1de61db1536727c1cad049c09decff22e8b6dd7:
io_uring/zcrx: document area chunking parameter (2026-01-14 02:13:37 +0000)
v9: - correct nits from Paolo
v8: - Add stripped down qcfg
- Retain the page size across resets for bnxt
v7: - Add xa_destroy
- Rebase
v6: - Update docs and add a selftest
v5: https://lore.kernel.org/netdev/[email protected]/
- Remove all unnecessary bits like configuration via netlink, and multi-stage queue configuration.
v4: https://lore.kernel.org/all/[email protected]/
- Update fbnic qops
- Propagate max buf len for hns3
- Use configured buf size in __bnxt_alloc_rx_netmem
- Minor stylistic changes
v3: https://lore.kernel.org/all/[email protected]/
- Rebased, excluded zcrx specific patches
- Set agg_size_fac to 1 on warning
v2: https://lore.kernel.org/all/[email protected]/
- Add MAX_PAGE_ORDER check on pp init
- Applied comments rewording
- Adjust pp.max_len based on order
- Patch up mlx5 queue callbacks after rebase
- Minor ->queue_mgmt_ops refactoring
- Rebased to account for both fill level and agg_size_fac
- Pass providers buf length in struct pp_memory_provider_params and apply it in __netdev_queue_confi().
- Use ->supported_ring_params to validate drivers support of set qcfg parameters.
Jakub Kicinski (2):
net: reduce indent of struct netdev_queue_mgmt_ops members
eth: bnxt: adjust the fill level of agg queues with larger buffers
Pavel Begunkov (7):
net: memzero mp params when closing a queue
net: add bare bone queue configs
net: pass queue rx page size from memory provider
eth: bnxt: store rx buffer size per queue
eth: bnxt: support qcfg provided rx page size
selftests: iou-zcrx: test large chunk sizes io_uring/zcrx: document area chunking parameter
Documentation/networking/iou-zcrx.rst | 20 +++ drivers/net/ethernet/broadcom/bnxt/bnxt.c | 123 +++++++++++++----- drivers/net/ethernet/broadcom/bnxt/bnxt.h | 2 + drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c | 6 +- drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.h | 2 +- drivers/net/ethernet/google/gve/gve_main.c | 9 +- .../net/ethernet/mellanox/mlx5/core/en_main.c | 10 +- drivers/net/ethernet/meta/fbnic/fbnic_txrx.c | 8 +- drivers/net/netdevsim/netdev.c | 7 +- include/net/netdev_queues.h | 47 +++++-- include/net/netdev_rx_queue.h | 2 + include/net/page_pool/types.h | 1 + net/core/dev.c | 17 +++ net/core/netdev_rx_queue.c | 31 +++-- .../selftests/drivers/net/hw/iou-zcrx.c | 72 ++++++++-- .../selftests/drivers/net/hw/iou-zcrx.py | 39 ++++++ 16 files changed, 317 insertions(+), 79 deletions(-)
Signature · 1 line
2.52.0
PATCH v9 1/9 net: memzero mp params when closing a queue
Instead of resetting memory provider parameters one by one in __net_mp_{open,close}_rxq, memzero the entire structure. It'll be used to extend the structure.
Signed-off-by Pavel Begunkov <[email protected]>
--- net/core/netdev_rx_queue.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-)
net/core/netdev_rx_queue.c +4 -6
@@ -139,10 +139,9 @@ int __net_mp_open_rxq(struct net_device *dev, unsigned int rxq_idx, rxq->mp_params = *p; ret =netdev_rx_queue_restart (dev, rxq_idx); -if (ret) { - rxq->mp_params.mp_ops =NULL ; - rxq->mp_params.mp_priv =NULL ; - } +if (ret) +memset (&rxq->mp_params,0 ,sizeof (rxq->mp_params)); +return ret; }
@@ -179,8 +178,7 @@ void __net_mp_close_rxq(struct net_device *dev, unsigned int ifq_idx, rxq->mp_params.mp_priv != old_p->mp_priv)) return; - rxq->mp_params.mp_ops =NULL ; - rxq->mp_params.mp_priv =NULL ; +memset (&rxq->mp_params,0 ,sizeof (rxq->mp_params)); err =netdev_rx_queue_restart (dev, ifq_idx);WARN_ON (err && err != -ENETDOWN ); }
Signature · 1 line
2.52.0
Re: [PATCH net-next v9 1/9] net: memzero mp params when closing a queue
This series was applied to netdev/net-next.git (main) by Pavel Begunkov <[email protected]>:
Hello:
This series was applied to netdev/net-next.git (main) by Pavel Begunkov <[email protected]>:
Quoted 8 lines · On Thu, 15 Jan 2026 17:11:54 +0000 you wrote:
Instead of resetting memory provider parameters one by one in __net_mp_{open,close}_rxq, memzero the entire structure. It'll be used to extend the structure.
Signed-off-by Pavel Begunkov <[email protected]>--- net/core/netdev_rx_queue.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-)
Here is the summary with links:
- [net-next,v9,1/9] net: memzero mp params when closing a queue
https://git.kernel.org/netdev/net-next/c/7073bb4b696f - [net-next,v9,2/9] net: reduce indent of struct netdev_queue_mgmt_ops members https://git.kernel.org/netdev/net-next/c/92d76cf96dcb - [net-next,v9,3/9] net: add bare bone queue configs https://git.kernel.org/netdev/net-next/c/efcb9a4d32d3 - [net-next,v9,4/9] net: pass queue rx page size from memory provider https://git.kernel.org/netdev/net-next/c/c0b709bf438b - [net-next,v9,5/9] eth: bnxt: store rx buffer size per queue https://git.kernel.org/netdev/net-next/c/f57efb32aae1 - [net-next,v9,6/9] eth: bnxt: adjust the fill level of agg queues with larger buffers https://git.kernel.org/netdev/net-next/c/c55bf90a2112 - [net-next,v9,7/9] eth: bnxt: support qcfg provided rx page size https://git.kernel.org/netdev/net-next/c/f96e1b35779e - [net-next,v9,8/9] selftests: iou-zcrx: test large chunk sizes https://git.kernel.org/netdev/net-next/c/a32bb32d0193 - [net-next,v9,9/9] io_uring/zcrx: document area chunking parameter https://git.kernel.org/netdev/net-next/c/d1de61db1536
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html
PATCH v9 2/9 net: reduce indent of struct netdev_queue_mgmt_ops members
In reply to Pavel BegunkovFrom: Jakub Kicinski <[email protected]>
Trivial change, reduce the indent. I think the original is copied from real NDOs. It's unnecessarily deep, makes passing struct args problematic.
Signed-off-by Jakub Kicinski <[email protected]>
Reviewed-by Mina Almasry <[email protected]>
Signed-off-by Pavel Begunkov <[email protected]>
--- include/net/netdev_queues.h | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-)
include/net/netdev_queues.h +14 -14
@@ -135,20 +135,20 @@ void netdev_stat_queue_sum(struct net_device *netdev,* be called for an interface which is open. */ struct netdev_queue_mgmt_ops { -size_t ndo_queue_mem_size; -int (*ndo_queue_mem_alloc)(struct net_device *dev, -void *per_queue_mem, -int idx); -void (*ndo_queue_mem_free)(struct net_device *dev, -void *per_queue_mem); -int (*ndo_queue_start)(struct net_device *dev, -void *per_queue_mem, -int idx); -int (*ndo_queue_stop)(struct net_device *dev, -void *per_queue_mem, -int idx); -struct device * (*ndo_queue_get_dma_dev)(struct net_device *dev, -int idx); +size_t ndo_queue_mem_size; +int (*ndo_queue_mem_alloc)(struct net_device *dev, +void *per_queue_mem, +int idx); +void (*ndo_queue_mem_free)(struct net_device *dev, +void *per_queue_mem); +int (*ndo_queue_start)(struct net_device *dev, +void *per_queue_mem, +int idx); +int (*ndo_queue_stop)(struct net_device *dev, +void *per_queue_mem, +int idx); +struct device * (*ndo_queue_get_dma_dev)(struct net_device *dev, +int idx); };bool netif_rxq_has_unreadable_mp (struct net_device *dev,int idx);
Signature · 1 line
2.52.0
PATCH v9 3/9 net: add bare bone queue configs
In reply to Pavel BegunkovWe'll need to pass extra parameters when allocating a queue for memory providers. Define a new structure for queue configurations, and pass it to qapi callbacks. It's empty for now, actual parameters will be added in following patches.
Configurations should persist across resets, and for that they're default-initialised on device registration and stored in struct netdev_rx_queue. We also add a new qapi callback for defaulting a given config. It must be implemented if a driver wants to use queue configs and is optional otherwise.
Suggested-by Jakub Kicinski <[email protected]>
Signed-off-by Pavel Begunkov <[email protected]>
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 8 ++++++-- drivers/net/ethernet/google/gve/gve_main.c | 9 ++++++--- .../net/ethernet/mellanox/mlx5/core/en_main.c | 10 ++++++---- drivers/net/ethernet/meta/fbnic/fbnic_txrx.c | 8 ++++++-- drivers/net/netdevsim/netdev.c | 7 +++++-- include/net/netdev_queues.h | 9 +++++++++ include/net/netdev_rx_queue.h | 2 ++ net/core/dev.c | 17 +++++++++++++++++ net/core/netdev_rx_queue.c | 12 +++++++++--- 9 files changed, 66 insertions(+), 16 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +6 -2
@@ -15911,7 +15911,9 @@ static const struct netdev_stat_ops bnxt_stat_ops = { .get_base_stats = bnxt_get_base_stats, }; -static int bnxt_queue_mem_alloc (struct net_device *dev,void *qmem,int idx) +static int bnxt_queue_mem_alloc (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *qmem,int idx) {struct bnxt_rx_ring_info *rxr, *clone;struct bnxt *bp =netdev_priv (dev);
@@ -16077,7 +16079,9 @@ static void bnxt_copy_rx_ring(struct bnxt *bp, dst->rx_agg_bmap = src->rx_agg_bmap; } -static int bnxt_queue_start (struct net_device *dev,void *qmem,int idx) +static int bnxt_queue_start (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *qmem,int idx) {struct bnxt *bp =netdev_priv (dev);struct bnxt_rx_ring_info *rxr, *clone;
drivers/net/ethernet/google/gve/gve_main.c +6 -3
@@ -2616,8 +2616,9 @@ static void gve_rx_queue_mem_free(struct net_device *dev, void *per_q_mem)gve_rx_free_ring_dqo (priv, gve_per_q_mem, &cfg); } -static int gve_rx_queue_mem_alloc (struct net_device *dev,void *per_q_mem, -int idx) +static int gve_rx_queue_mem_alloc (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *per_q_mem,int idx) {struct gve_priv *priv =netdev_priv (dev);struct gve_rx_alloc_rings_cfg cfg = {0 };
@@ -2638,7 +2639,9 @@ static int gve_rx_queue_mem_alloc(struct net_device *dev, void *per_q_mem,return err; } -static int gve_rx_queue_start (struct net_device *dev,void *per_q_mem,int idx) +static int gve_rx_queue_start (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *per_q_mem,int idx) {struct gve_priv *priv =netdev_priv (dev);struct gve_rx_ring *gve_per_q_mem;
drivers/net/ethernet/mellanox/mlx5/core/en_main.c +6 -4
@@ -5596,8 +5596,9 @@ struct mlx5_qmgmt_data {struct mlx5e_channel_param cparam; }; -static int mlx5e_queue_mem_alloc (struct net_device *dev,void *newq, -int queue_index) +static int mlx5e_queue_mem_alloc (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *newq,int queue_index) {struct mlx5_qmgmt_data *new = (struct mlx5_qmgmt_data *)newq;struct mlx5e_priv *priv =netdev_priv (dev);
@@ -5658,8 +5659,9 @@ static int mlx5e_queue_stop(struct net_device *dev, void *oldq, int queue_index)return 0 ; } -static int mlx5e_queue_start (struct net_device *dev,void *newq, -int queue_index) +static int mlx5e_queue_start (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *newq,int queue_index) {struct mlx5_qmgmt_data *new = (struct mlx5_qmgmt_data *)newq;struct mlx5e_priv *priv =netdev_priv (dev);
drivers/net/ethernet/meta/fbnic/fbnic_txrx.c +6 -2
@@ -2809,7 +2809,9 @@ void fbnic_napi_depletion_check(struct net_device *netdev)fbnic_wrfl (fbd); } -static int fbnic_queue_mem_alloc (struct net_device *dev,void *qmem,int idx) +static int fbnic_queue_mem_alloc (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *qmem,int idx) {struct fbnic_net *fbn =netdev_priv (dev);const struct fbnic_q_triad *real;
@@ -2861,7 +2863,9 @@ static void __fbnic_nv_restart(struct fbnic_net *fbn,netif_wake_subqueue (fbn->netdev, nv->qt[i].sub0.q_idx); } -static int fbnic_queue_start (struct net_device *dev,void *qmem,int idx) +static int fbnic_queue_start (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *qmem,int idx) {struct fbnic_net *fbn =netdev_priv (dev);struct fbnic_napi_vector *nv;
drivers/net/netdevsim/netdev.c +5 -2
@@ -758,7 +758,9 @@ struct nsim_queue_mem { };static int -nsim_queue_mem_alloc (struct net_device *dev,void *per_queue_mem,int idx) +nsim_queue_mem_alloc (struct net_device *dev, +struct netdev_queue_config *qcfg, +void *per_queue_mem,int idx) {struct nsim_queue_mem *qmem = per_queue_mem;struct netdevsim *ns =netdev_priv (dev);
@@ -807,7 +809,8 @@ static void nsim_queue_mem_free(struct net_device *dev, void *per_queue_mem) }static int -nsim_queue_start (struct net_device *dev,void *per_queue_mem,int idx) +nsim_queue_start (struct net_device *dev,struct netdev_queue_config *qcfg, +void *per_queue_mem,int idx) {struct nsim_queue_mem *qmem = per_queue_mem;struct netdevsim *ns =netdev_priv (dev);
include/net/netdev_queues.h +9 -0
@@ -14,6 +14,9 @@ struct netdev_config {u8 hds_config; }; +struct netdev_queue_config { +}; +/* See the netdev.yaml spec for definition of each statistic */ struct netdev_queue_stats_rx {u64 bytes;
@@ -130,6 +133,8 @@ void netdev_stat_queue_sum(struct net_device *netdev,* @ndo_queue_get_dma_dev: Get dma device for zero-copy operations to be used * for this queue. Return NULL on error. * +* @ndo_default_qcfg: Populate queue config struct with defaults. Optional. +* * Note that @ndo_queue_mem_alloc and @ndo_queue_mem_free may be called while * the interface is closed. @ndo_queue_start and @ndo_queue_stop will only * be called for an interface which is open.
@@ -137,16 +142,20 @@ void netdev_stat_queue_sum(struct net_device *netdev,struct netdev_queue_mgmt_ops {size_t ndo_queue_mem_size;int (*ndo_queue_mem_alloc)(struct net_device *dev, +struct netdev_queue_config *qcfg,void *per_queue_mem,int idx);void (*ndo_queue_mem_free)(struct net_device *dev,void *per_queue_mem);int (*ndo_queue_start)(struct net_device *dev, +struct netdev_queue_config *qcfg,void *per_queue_mem,int idx);int (*ndo_queue_stop)(struct net_device *dev,void *per_queue_mem,int idx); +void (*ndo_default_qcfg)(struct net_device *dev, +struct netdev_queue_config *qcfg);struct device * (*ndo_queue_get_dma_dev)(struct net_device *dev,int idx); };
include/net/netdev_rx_queue.h +2 -0
@@ -7,6 +7,7 @@#include <linux/sysfs.h> #include <net/xdp.h> #include <net/page_pool/types.h> +#include <net/netdev_queues.h> /* This structure contains an instance of an RX queue. */ struct netdev_rx_queue {
@@ -27,6 +28,7 @@ struct netdev_rx_queue {struct xsk_buff_pool *pool;#endif struct napi_struct *napi; +struct netdev_queue_config qcfg;struct pp_memory_provider_params mp_params; }____cacheline_aligned_in_smp ;
net/core/dev.c +17 -0
@@ -11270,6 +11270,21 @@ static void netdev_free_phy_link_topology(struct net_device *dev) } } +static void init_rx_queue_cfgs (struct net_device *dev) +{ +const struct netdev_queue_mgmt_ops *qops = dev->queue_mgmt_ops; +struct netdev_rx_queue *rxq; +int i; + +if (!qops || !qops->ndo_default_qcfg) +return ; + +for (i =0 ; i < dev->num_rx_queues; i++) { + rxq =__netif_get_rx_queue (dev, i); + qops->ndo_default_qcfg (dev, &rxq->qcfg); + } +} +/** * register_netdevice() - register a network device * @dev: device to register
@@ -11315,6 +11330,8 @@ int register_netdevice(struct net_device *dev)if (!dev->name_node) goto out; +init_rx_queue_cfgs (dev); +/* Init, if this function is available */ if (dev->netdev_ops->ndo_init) { ret = dev->netdev_ops->ndo_init (dev);
net/core/netdev_rx_queue.c +9 -3
@@ -22,6 +22,7 @@ int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq_idx) {struct netdev_rx_queue *rxq =__netif_get_rx_queue (dev, rxq_idx);const struct netdev_queue_mgmt_ops *qops = dev->queue_mgmt_ops; +struct netdev_queue_config qcfg;void *new_mem, *old_mem;int err;
@@ -31,6 +32,10 @@ int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq_idx)netdev_assert_locked (dev); +memset (&qcfg,0 ,sizeof (qcfg)); +if (qops->ndo_default_qcfg) + qops->ndo_default_qcfg (dev, &qcfg); + new_mem =kvzalloc (qops->ndo_queue_mem_size,GFP_KERNEL );if (!new_mem)return -ENOMEM ;
@@ -41,7 +46,7 @@ int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq_idx) goto err_free_new_mem; } - err = qops->ndo_queue_mem_alloc (dev, new_mem, rxq_idx); + err = qops->ndo_queue_mem_alloc (dev, &qcfg, new_mem, rxq_idx);if (err) goto err_free_old_mem;
@@ -54,7 +59,7 @@ int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq_idx)if (err) goto err_free_new_queue_mem; - err = qops->ndo_queue_start (dev, new_mem, rxq_idx); + err = qops->ndo_queue_start (dev, &qcfg, new_mem, rxq_idx);if (err) goto err_start_queue; }else {
@@ -66,6 +71,7 @@ int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq_idx)kvfree (old_mem);kvfree (new_mem); + rxq->qcfg = qcfg;return 0 ; err_start_queue:
@@ -76,7 +82,7 @@ int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq_idx)* WARN if we fail to recover the old rx queue, and at least free * old_mem so we don't also leak that. */ -if (qops->ndo_queue_start (dev, old_mem, rxq_idx)) { +if (qops->ndo_queue_start (dev, &rxq->qcfg, old_mem, rxq_idx)) {WARN (1 ,"Failed to restart old queue in error path. RX queue %d may be unhealthy." , rxq_idx);
Signature · 1 line
2.52.0
PATCH v9 4/9 net: pass queue rx page size from memory provider
In reply to Pavel BegunkovAllow memory providers to configure rx queues with a custom receive page size. It's passed in struct pp_memory_provider_params, which is copied into the queue, so it's preserved across queue restarts. Then, it's propagated to the driver in a new queue config parameter.
Drivers should explicitly opt into using it by setting QCFG_RX_PAGE_SIZE, in which case they should implement ndo_default_qcfg, validate the size on queue restart and honour the current config in case of a reset.
Signed-off-by Pavel Begunkov <[email protected]>
--- include/net/netdev_queues.h | 10 ++++++++++ include/net/page_pool/types.h | 1 + net/core/netdev_rx_queue.c | 9 +++++++++ 3 files changed, 20 insertions(+)
include/net/netdev_queues.h +10 -0
@@ -15,6 +15,7 @@ struct netdev_config { };struct netdev_queue_config { +u32 rx_page_size; };/* See the netdev.yaml spec for definition of each statistic */
@@ -114,6 +115,11 @@ void netdev_stat_queue_sum(struct net_device *netdev, int tx_start, int tx_end, struct netdev_queue_stats_tx *tx_sum); +enum { +/* The queue checks and honours the page size qcfg parameter */ +QCFG_RX_PAGE_SIZE =0x1 , +}; +/** * struct netdev_queue_mgmt_ops - netdev ops for queue management *
@@ -135,6 +141,8 @@ void netdev_stat_queue_sum(struct net_device *netdev,* * @ndo_default_qcfg: Populate queue config struct with defaults. Optional. * +* @supported_params: Bitmask of supported parameters, see QCFG_*. +* * Note that @ndo_queue_mem_alloc and @ndo_queue_mem_free may be called while * the interface is closed. @ndo_queue_start and @ndo_queue_stop will only * be called for an interface which is open.
@@ -158,6 +166,8 @@ struct netdev_queue_mgmt_ops { struct netdev_queue_config *qcfg);struct device * (*ndo_queue_get_dma_dev)(struct net_device *dev,int idx); + +unsigned int supported_params; };bool netif_rxq_has_unreadable_mp (struct net_device *dev,int idx);
include/net/page_pool/types.h +1 -0
@@ -161,6 +161,7 @@ struct memory_provider_ops;struct pp_memory_provider_params {void *mp_priv;const struct memory_provider_ops *mp_ops; +u32 rx_page_size; };struct page_pool {
net/core/netdev_rx_queue.c +9 -0
@@ -30,12 +30,21 @@ int netdev_rx_queue_restart(struct net_device *dev, unsigned int rxq_idx) !qops->ndo_queue_mem_alloc || !qops->ndo_queue_start)return -EOPNOTSUPP ; +if (WARN_ON_ONCE (qops->supported_params && !qops->ndo_default_qcfg)) +return -EINVAL ; +netdev_assert_locked (dev);memset (&qcfg,0 ,sizeof (qcfg));if (qops->ndo_default_qcfg) qops->ndo_default_qcfg (dev, &qcfg); +if (rxq->mp_params.rx_page_size) { +if (!(qops->supported_params &QCFG_RX_PAGE_SIZE )) +return -EOPNOTSUPP ; + qcfg.rx_page_size = rxq->mp_params.rx_page_size; + } + new_mem =kvzalloc (qops->ndo_queue_mem_size,GFP_KERNEL );if (!new_mem)return -ENOMEM ;
Signature · 1 line
2.52.0
PATCH v9 5/9 eth: bnxt: store rx buffer size per queue
In reply to Pavel BegunkovInstead of using a constant buffer length, allow configuring the size for each queue separately. There is no way to change the length yet, and it'll be passed from memory providers in a later patch.
Suggested-by Jakub Kicinski <[email protected]>
Signed-off-by Pavel Begunkov <[email protected]>
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 56 +++++++++++-------- drivers/net/ethernet/broadcom/bnxt/bnxt.h | 1 + drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c | 6 +- drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.h | 2 +- 4 files changed, 38 insertions(+), 27 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +33 -23
@@ -905,7 +905,7 @@ static void bnxt_tx_int(struct bnxt *bp, struct bnxt_napi *bnapi, int budget)static bool bnxt_separate_head_pool (struct bnxt_rx_ring_info *rxr) { -return rxr->need_head_pool ||PAGE_SIZE >BNXT_RX_PAGE_SIZE ; +return rxr->need_head_pool || rxr->rx_page_size <PAGE_SIZE ; } static struct page *__bnxt_alloc_rx_page(struct bnxt *bp, dma_addr_t *mapping,
@@ -915,9 +915,9 @@ static struct page *__bnxt_alloc_rx_page(struct bnxt *bp, dma_addr_t *mapping, {struct page *page; -if (PAGE_SIZE >BNXT_RX_PAGE_SIZE ) { +if (rxr->rx_page_size <PAGE_SIZE ) { page =page_pool_dev_alloc_frag (rxr->page_pool, offset, -BNXT_RX_PAGE_SIZE ); + rxr->rx_page_size); }else { page =page_pool_dev_alloc_pages (rxr->page_pool); *offset =0 ;
@@ -936,8 +936,9 @@ static netmem_ref __bnxt_alloc_rx_netmem(struct bnxt *bp, dma_addr_t *mapping, {netmem_ref netmem; -if (PAGE_SIZE >BNXT_RX_PAGE_SIZE ) { - netmem =page_pool_alloc_frag_netmem (rxr->page_pool, offset,BNXT_RX_PAGE_SIZE , gfp); +if (rxr->rx_page_size <PAGE_SIZE ) { + netmem =page_pool_alloc_frag_netmem (rxr->page_pool, offset, + rxr->rx_page_size, gfp); }else { netmem =page_pool_alloc_netmems (rxr->page_pool, gfp); *offset =0 ;
@@ -1155,9 +1156,9 @@ static struct sk_buff *bnxt_rx_multi_page_skb(struct bnxt *bp,return NULL ; } dma_addr -= bp->rx_dma_offset; -dma_sync_single_for_cpu (&bp->pdev->dev, dma_addr,BNXT_RX_PAGE_SIZE , +dma_sync_single_for_cpu (&bp->pdev->dev, dma_addr, rxr->rx_page_size, bp->rx_dir); - skb =napi_build_skb (data_ptr - bp->rx_offset,BNXT_RX_PAGE_SIZE ); + skb =napi_build_skb (data_ptr - bp->rx_offset, rxr->rx_page_size);if (!skb) {page_pool_recycle_direct (rxr->page_pool, page);return NULL ;
@@ -1189,7 +1190,7 @@ static struct sk_buff *bnxt_rx_page_skb(struct bnxt *bp,return NULL ; } dma_addr -= bp->rx_dma_offset; -dma_sync_single_for_cpu (&bp->pdev->dev, dma_addr,BNXT_RX_PAGE_SIZE , +dma_sync_single_for_cpu (&bp->pdev->dev, dma_addr, rxr->rx_page_size, bp->rx_dir);if (unlikely (!payload))
@@ -1203,7 +1204,7 @@ static struct sk_buff *bnxt_rx_page_skb(struct bnxt *bp,skb_mark_for_recycle (skb); off = (void *)data_ptr -page_address (page); -skb_add_rx_frag (skb,0 , page, off, len,BNXT_RX_PAGE_SIZE ); +skb_add_rx_frag (skb,0 , page, off, len, rxr->rx_page_size);memcpy (skb->data -NET_IP_ALIGN , data_ptr -NET_IP_ALIGN , payload +NET_IP_ALIGN );
@@ -1288,7 +1289,7 @@ static u32 __bnxt_rx_agg_netmems(struct bnxt *bp,if (skb) {skb_add_rx_frag_netmem (skb, i, cons_rx_buf->netmem, cons_rx_buf->offset, - frag_len,BNXT_RX_PAGE_SIZE ); + frag_len, rxr->rx_page_size); }else {skb_frag_t *frag = &shinfo->frags[i];
@@ -1313,7 +1314,7 @@ static u32 __bnxt_rx_agg_netmems(struct bnxt *bp,if (skb) { skb->len -= frag_len; skb->data_len -= frag_len; - skb->truesize -=BNXT_RX_PAGE_SIZE ; + skb->truesize -= rxr->rx_page_size; } --shinfo->nr_frags;
@@ -1328,7 +1329,7 @@ static u32 __bnxt_rx_agg_netmems(struct bnxt *bp, }page_pool_dma_sync_netmem_for_cpu (rxr->page_pool, netmem,0 , -BNXT_RX_PAGE_SIZE ); + rxr->rx_page_size); total_frag_len += frag_len; prod =NEXT_RX_AGG (prod);
@@ -2290,8 +2291,7 @@ static int bnxt_rx_pkt(struct bnxt *bp, struct bnxt_cp_ring_info *cpr,if (!skb) goto oom_next_rx; }else { - skb =bnxt_xdp_build_skb (bp, skb, agg_bufs, - rxr->page_pool, &xdp); + skb =bnxt_xdp_build_skb (bp, skb, agg_bufs, rxr, &xdp);if (!skb) {/* we should be able to free the old skb here */ bnxt_xdp_buff_frags_free (rxr, &xdp);
@@ -3837,11 +3837,13 @@ static int bnxt_alloc_rx_page_pool(struct bnxt *bp, pp.pool_size = bp->rx_agg_ring_size / agg_size_fac;if (BNXT_RX_PAGE_MODE (bp)) pp.pool_size += bp->rx_ring_size / rx_size_fac; + + pp.order =get_order (rxr->rx_page_size); pp.nid = numa_node; pp.netdev = bp->dev; pp.dev = &bp->pdev->dev; pp.dma_dir = bp->rx_dir; - pp.max_len =PAGE_SIZE ; + pp.max_len =PAGE_SIZE << pp.order; pp.flags =PP_FLAG_DMA_MAP |PP_FLAG_DMA_SYNC_DEV |PP_FLAG_ALLOW_UNREADABLE_NETMEM ; pp.queue_idx = rxr->bnapi->index;
@@ -3852,7 +3854,10 @@ static int bnxt_alloc_rx_page_pool(struct bnxt *bp, rxr->page_pool = pool; rxr->need_head_pool =page_pool_is_unreadable (pool); + rxr->need_head_pool |= !!pp.order;if (bnxt_separate_head_pool (rxr)) { + pp.order =0 ; + pp.max_len =PAGE_SIZE ; pp.pool_size =min (bp->rx_ring_size / rx_size_fac,1024 ); pp.flags =PP_FLAG_DMA_MAP |PP_FLAG_DMA_SYNC_DEV ; pool =page_pool_create (&pp);
@@ -4328,6 +4333,8 @@ static void bnxt_init_ring_struct(struct bnxt *bp)if (!rxr) goto skip_rx; + rxr->rx_page_size =BNXT_RX_PAGE_SIZE ; + ring = &rxr->rx_ring_struct; rmem = &ring->ring_mem; rmem->nr_pages = bp->rx_nr_pages;
@@ -4487,7 +4494,7 @@ static void bnxt_init_one_rx_agg_ring_rxbd(struct bnxt *bp, ring = &rxr->rx_agg_ring_struct; ring->fw_ring_id =INVALID_HW_RING_ID ;if ((bp->flags &BNXT_FLAG_AGG_RINGS )) { - type = ((u32 )BNXT_RX_PAGE_SIZE <<RX_BD_LEN_SHIFT ) | + type = ((u32 )rxr->rx_page_size <<RX_BD_LEN_SHIFT ) |RX_BD_TYPE_RX_AGG_BD ;/* On P7, setting EOP will cause the chip to disable
@@ -7065,6 +7072,7 @@ static void bnxt_hwrm_ring_grp_free(struct bnxt *bp)static void bnxt_set_rx_ring_params_p5 (struct bnxt *bp,u32 ring_type,struct hwrm_ring_alloc_input *req, +struct bnxt_rx_ring_info *rxr,struct bnxt_ring_struct *ring) {struct bnxt_ring_grp_info *grp_info = &bp->grp_info[ring->grp_idx];
@@ -7074,7 +7082,7 @@ static void bnxt_set_rx_ring_params_p5(struct bnxt *bp, u32 ring_type,if (ring_type ==HWRM_RING_ALLOC_AGG ) { req->ring_type =RING_ALLOC_REQ_RING_TYPE_RX_AGG ; req->rx_ring_id =cpu_to_le16 (grp_info->rx_fw_ring_id); - req->rx_buf_size =cpu_to_le16 (BNXT_RX_PAGE_SIZE ); + req->rx_buf_size =cpu_to_le16 (rxr->rx_page_size); enables |=RING_ALLOC_REQ_ENABLES_RX_RING_ID_VALID ; }else { req->rx_buf_size =cpu_to_le16 (bp->rx_buf_use_size);
@@ -7088,6 +7096,7 @@ static void bnxt_set_rx_ring_params_p5(struct bnxt *bp, u32 ring_type, }static int hwrm_ring_alloc_send_msg (struct bnxt *bp, +struct bnxt_rx_ring_info *rxr,struct bnxt_ring_struct *ring,u32 ring_type,u32 map_index) {
@@ -7144,7 +7153,8 @@ static int hwrm_ring_alloc_send_msg(struct bnxt *bp, cpu_to_le32(bp->rx_ring_mask + 1) :cpu_to_le32 (bp->rx_agg_ring_mask +1 );if (bp->flags &BNXT_FLAG_CHIP_P5_PLUS ) -bnxt_set_rx_ring_params_p5 (bp, ring_type, req, ring); +bnxt_set_rx_ring_params_p5 (bp, ring_type, req, + rxr, ring);break ;case HWRM_RING_ALLOC_CMPL : req->ring_type =RING_ALLOC_REQ_RING_TYPE_L2_CMPL ;
@@ -7292,7 +7302,7 @@ static int bnxt_hwrm_rx_ring_alloc(struct bnxt *bp,u32 map_idx = bnapi->index;int rc; - rc =hwrm_ring_alloc_send_msg (bp, ring, type, map_idx); + rc =hwrm_ring_alloc_send_msg (bp, rxr, ring, type, map_idx);if (rc)return rc;
@@ -7312,7 +7322,7 @@ static int bnxt_hwrm_rx_agg_ring_alloc(struct bnxt *bp,int rc; map_idx = grp_idx + bp->rx_nr_rings; - rc =hwrm_ring_alloc_send_msg (bp, ring, type, map_idx); + rc =hwrm_ring_alloc_send_msg (bp, rxr, ring, type, map_idx);if (rc)return rc;
@@ -7336,7 +7346,7 @@ static int bnxt_hwrm_cp_ring_alloc_p5(struct bnxt *bp, ring = &cpr->cp_ring_struct; ring->handle =BNXT_SET_NQ_HDL (cpr); - rc =hwrm_ring_alloc_send_msg (bp, ring, type, map_idx); + rc =hwrm_ring_alloc_send_msg (bp,NULL , ring, type, map_idx);if (rc)return rc;bnxt_set_db (bp, &cpr->cp_db, type, map_idx, ring->fw_ring_id);
@@ -7351,7 +7361,7 @@ static int bnxt_hwrm_tx_ring_alloc(struct bnxt *bp,const u32 type =HWRM_RING_ALLOC_TX ;int rc; - rc =hwrm_ring_alloc_send_msg (bp, ring, type, tx_idx); + rc =hwrm_ring_alloc_send_msg (bp,NULL , ring, type, tx_idx);if (rc)return rc;bnxt_set_db (bp, &txr->tx_db, type, tx_idx, ring->fw_ring_id);
@@ -7377,7 +7387,7 @@ static int bnxt_hwrm_ring_alloc(struct bnxt *bp) vector = bp->irq_tbl[map_idx].vector;disable_irq_nosync (vector); - rc =hwrm_ring_alloc_send_msg (bp, ring, type, map_idx); + rc =hwrm_ring_alloc_send_msg (bp,NULL , ring, type, map_idx);if (rc) {enable_irq (vector); goto err_out;
drivers/net/ethernet/broadcom/bnxt/bnxt.h +1 -0
@@ -1105,6 +1105,7 @@ struct bnxt_rx_ring_info {unsigned long *rx_agg_bmap;u16 rx_agg_bmap_size; +u32 rx_page_size;bool need_head_pool;dma_addr_t rx_desc_mapping[MAX_RX_PAGES ];
drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.c +3 -3
@@ -183,7 +183,7 @@ void bnxt_xdp_buff_init(struct bnxt *bp, struct bnxt_rx_ring_info *rxr, u16 cons, u8 *data_ptr, unsigned int len, struct xdp_buff *xdp) { - u32 buflen = BNXT_RX_PAGE_SIZE; + u32 buflen = rxr->rx_page_size;struct bnxt_sw_rx_bd *rx_buf;struct pci_dev *pdev;dma_addr_t mapping;
@@ -460,7 +460,7 @@ int bnxt_xdp(struct net_device *dev, struct netdev_bpf *xdp)struct sk_buff *bnxt_xdp_build_skb (struct bnxt *bp,struct sk_buff *skb,u8 num_frags, -struct page_pool *pool,struct xdp_buff *xdp) +struct bnxt_rx_ring_info *rxr,struct xdp_buff *xdp) {struct skb_shared_info *sinfo =xdp_get_shared_info_from_buff (xdp);
@@ -468,7 +468,7 @@ bnxt_xdp_build_skb(struct bnxt *bp, struct sk_buff *skb, u8 num_frags,return NULL ;xdp_update_skb_frags_info (skb, num_frags, sinfo->xdp_frags_size, -BNXT_RX_PAGE_SIZE * num_frags, + rxr->rx_page_size * num_frags,xdp_buff_get_skb_flags (xdp));return skb; }
drivers/net/ethernet/broadcom/bnxt/bnxt_xdp.h +1 -1
@@ -32,6 +32,6 @@ void bnxt_xdp_buff_init(struct bnxt *bp, struct bnxt_rx_ring_info *rxr,void bnxt_xdp_buff_frags_free (struct bnxt_rx_ring_info *rxr,struct xdp_buff *xdp);struct sk_buff *bnxt_xdp_build_skb (struct bnxt *bp,struct sk_buff *skb, -u8 num_frags,struct page_pool *pool, +u8 num_frags,struct bnxt_rx_ring_info *rxr,struct xdp_buff *xdp);#endif
Signature · 1 line
2.52.0
PATCH v9 6/9 eth: bnxt: adjust the fill level of agg queues with larger buffers
In reply to Pavel BegunkovFrom: Jakub Kicinski <[email protected]>
The driver tries to provision more agg buffers than header buffers since multiple agg segments can reuse the same header. The calculation / heuristic tries to provide enough pages for 65k of data for each header (or 4 frags per header if the result is too big). This calculation is currently global to the adapter. If we increase the buffer sizes 8x we don't want 8x the amount of memory sitting on the rings. Luckily we don't have to fill the rings completely, adjust the fill level dynamically in case particular queue has buffers larger than the global size.
Signed-off-by Jakub Kicinski <[email protected]>
[pavel: rebase on top of agg_size_fac]
Signed-off-by Pavel Begunkov <[email protected]>
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 25 +++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +21 -4
@@ -3825,16 +3825,31 @@ static void bnxt_free_rx_rings(struct bnxt *bp) } } +static int bnxt_rx_agg_ring_fill_level (struct bnxt *bp, +struct bnxt_rx_ring_info *rxr) +{ +/* User may have chosen larger than default rx_page_size, +* we keep the ring sizes uniform and also want uniform amount +* of bytes consumed per ring, so cap how much of the rings we fill. +*/ +int fill_level = bp->rx_agg_ring_size; + +if (rxr->rx_page_size >BNXT_RX_PAGE_SIZE ) + fill_level /= rxr->rx_page_size /BNXT_RX_PAGE_SIZE ; + +return fill_level; +} +static int bnxt_alloc_rx_page_pool (struct bnxt *bp,struct bnxt_rx_ring_info *rxr,int numa_node) { -const unsigned int agg_size_fac =PAGE_SIZE /BNXT_RX_PAGE_SIZE ; +unsigned int agg_size_fac = rxr->rx_page_size /BNXT_RX_PAGE_SIZE ;const unsigned int rx_size_fac =PAGE_SIZE /SZ_4K ;struct page_pool_params pp = {0 };struct page_pool *pool; - pp.pool_size = bp->rx_agg_ring_size / agg_size_fac; + pp.pool_size =bnxt_rx_agg_ring_fill_level (bp, rxr) / agg_size_fac;if (BNXT_RX_PAGE_MODE (bp)) pp.pool_size += bp->rx_ring_size / rx_size_fac;
@@ -4412,11 +4427,13 @@ static void bnxt_alloc_one_rx_ring_netmem(struct bnxt *bp,struct bnxt_rx_ring_info *rxr, int ring_nr) { + int fill_level, i;u32 prod; -int i; + + fill_level =bnxt_rx_agg_ring_fill_level (bp, rxr); prod = rxr->rx_agg_prod; -for (i =0 ; i < bp->rx_agg_ring_size; i++) { +for (i =0 ; i < fill_level; i++) {if (bnxt_alloc_rx_netmem (bp, rxr, prod,GFP_KERNEL )) {netdev_warn (bp->dev,"init'ed rx ring %d with %d/%d pages only\n" , ring_nr, i, bp->rx_agg_ring_size);
Signature · 1 line
2.52.0
PATCH v9 7/9 eth: bnxt: support qcfg provided rx page size
In reply to Pavel BegunkovImplement support for qcfg provided rx page sizes. For that, implement the ndo_default_qcfg callback and validate the config on restart. Also, use the current config's value in bnxt_init_ring_struct to retain the correct size across resets.
Signed-off-by Pavel Begunkov <[email protected]>
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 36 ++++++++++++++++++++++- drivers/net/ethernet/broadcom/bnxt/bnxt.h | 1 + 2 files changed, 36 insertions(+), 1 deletion(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +35 -1
@@ -4331,6 +4331,7 @@ static void bnxt_init_ring_struct(struct bnxt *bp)struct bnxt_rx_ring_info *rxr;struct bnxt_tx_ring_info *txr;struct bnxt_ring_struct *ring; +struct netdev_rx_queue *rxq;if (!bnapi)continue ;
@@ -4348,7 +4349,8 @@ static void bnxt_init_ring_struct(struct bnxt *bp)if (!rxr) goto skip_rx; - rxr->rx_page_size =BNXT_RX_PAGE_SIZE ; + rxq =__netif_get_rx_queue (bp->dev, i); + rxr->rx_page_size = rxq->qcfg.rx_page_size; ring = &rxr->rx_ring_struct; rmem = &ring->ring_mem;
@@ -15938,6 +15940,29 @@ static const struct netdev_stat_ops bnxt_stat_ops = { .get_base_stats = bnxt_get_base_stats, }; +static void bnxt_queue_default_qcfg (struct net_device *dev, +struct netdev_queue_config *qcfg) +{ + qcfg->rx_page_size =BNXT_RX_PAGE_SIZE ; +} + +static int bnxt_validate_qcfg (struct bnxt *bp,struct netdev_queue_config *qcfg) +{ +/* Older chips need MSS calc so rx_page_size is not supported */ +if (!(bp->flags &BNXT_FLAG_CHIP_P5_PLUS ) && + qcfg->rx_page_size !=BNXT_RX_PAGE_SIZE ) +return -EINVAL ; + +if (!is_power_of_2 (qcfg->rx_page_size)) +return -ERANGE ; + +if (qcfg->rx_page_size <BNXT_RX_PAGE_SIZE || + qcfg->rx_page_size >BNXT_MAX_RX_PAGE_SIZE ) +return -ERANGE ; + +return 0 ; +} +static int bnxt_queue_mem_alloc (struct net_device *dev,struct netdev_queue_config *qcfg,void *qmem,int idx)
@@ -15950,6 +15975,10 @@ static int bnxt_queue_mem_alloc(struct net_device *dev,if (!bp->rx_ring)return -ENETDOWN ; + rc =bnxt_validate_qcfg (bp, qcfg); +if (rc <0 ) +return rc; + rxr = &bp->rx_ring[idx]; clone = qmem;memcpy (clone, rxr,sizeof (*rxr));
@@ -15961,6 +15990,7 @@ static int bnxt_queue_mem_alloc(struct net_device *dev, clone->rx_sw_agg_prod =0 ; clone->rx_next_cons =0 ; clone->need_head_pool = false; + clone->rx_page_size = qcfg->rx_page_size; rc =bnxt_alloc_rx_page_pool (bp, clone, rxr->page_pool->p.nid);if (rc)
@@ -16087,6 +16117,8 @@ static void bnxt_copy_rx_ring(struct bnxt *bp, src_ring = &src->rx_agg_ring_struct; src_rmem = &src_ring->ring_mem; + dst->rx_page_size = src->rx_page_size; +WARN_ON (dst_rmem->nr_pages != src_rmem->nr_pages);WARN_ON (dst_rmem->page_size != src_rmem->page_size);WARN_ON (dst_rmem->flags != src_rmem->flags);
@@ -16241,6 +16273,8 @@ static const struct netdev_queue_mgmt_ops bnxt_queue_mgmt_ops = { .ndo_queue_mem_free = bnxt_queue_mem_free, .ndo_queue_start = bnxt_queue_start, .ndo_queue_stop = bnxt_queue_stop, + .ndo_default_qcfg = bnxt_queue_default_qcfg, + .supported_params = QCFG_RX_PAGE_SIZE, };static void bnxt_remove_one (struct pci_dev *pdev)
drivers/net/ethernet/broadcom/bnxt/bnxt.h +1 -0
@@ -760,6 +760,7 @@ struct nqe_cn {#endif #define BNXT_RX_PAGE_SIZE (1 << BNXT_RX_PAGE_SHIFT) +#define BNXT_MAX_RX_PAGE_SIZE BIT(15)#define BNXT_MAX_MTU 9500
Signature · 1 line
2.52.0
PATCH v9 8/9 selftests: iou-zcrx: test large chunk sizes
In reply to Pavel BegunkovAdd a test using large chunks for zcrx memory area.
Signed-off-by Pavel Begunkov <[email protected]>
--- .../selftests/drivers/net/hw/iou-zcrx.c | 72 +++++++++++++++---- .../selftests/drivers/net/hw/iou-zcrx.py | 39 ++++++++++ 2 files changed, 99 insertions(+), 12 deletions(-)
tools/testing/selftests/drivers/net/hw/iou-zcrx.c +60 -12
@@ -12,6 +12,7 @@#include <unistd.h> #include <arpa/inet.h> +#include <linux/mman.h> #include <linux/errqueue.h> #include <linux/if_packet.h> #include <linux/ipv6.h>
@@ -37,6 +38,23 @@#include <liburing.h> +#define SKIP_CODE 42 + +struct t_io_uring_zcrx_ifq_reg { +__u32 if_idx; +__u32 if_rxq; +__u32 rq_entries; +__u32 flags; + +__u64 area_ptr;/* pointer to struct io_uring_zcrx_area_reg */ +__u64 region_ptr;/* struct io_uring_region_desc * */ + +struct io_uring_zcrx_offsets offsets; +__u32 zcrx_id; +__u32 rx_buf_len; +__u64 __resv[3 ]; +}; +static long page_size;#define AREA_SIZE (8192 * page_size)#define SEND_SIZE (512 * 4096)
@@ -65,6 +83,8 @@ static bool cfg_oneshot;static int cfg_oneshot_recvs;static int cfg_send_size =SEND_SIZE ;static struct sockaddr_in6 cfg_addr; +static unsigned int cfg_rx_buf_len; +static bool cfg_dry_run;static char *payload;static void *area_ptr;
@@ -128,14 +148,28 @@ static void setup_zcrx(struct io_uring *ring)if (!ifindex)error (1 ,0 ,"bad interface name: %s" , cfg_ifname); - area_ptr =mmap (NULL , -AREA_SIZE , -PROT_READ |PROT_WRITE , -MAP_ANONYMOUS |MAP_PRIVATE , -0 , -0 ); -if (area_ptr ==MAP_FAILED ) -error (1 ,0 ,"mmap(): zero copy area" ); +if (cfg_rx_buf_len && cfg_rx_buf_len != page_size) { + area_ptr =mmap (NULL , +AREA_SIZE , +PROT_READ |PROT_WRITE , +MAP_ANONYMOUS |MAP_PRIVATE | +MAP_HUGETLB |MAP_HUGE_2MB , +-1 , +0 ); +if (area_ptr ==MAP_FAILED ) { +printf ("Can't allocate huge pages\n" ); +exit (SKIP_CODE ); + } + }else { + area_ptr =mmap (NULL , +AREA_SIZE , +PROT_READ |PROT_WRITE , +MAP_ANONYMOUS |MAP_PRIVATE , +0 , +0 ); +if (area_ptr ==MAP_FAILED ) +error (1 ,0 ,"mmap(): zero copy area" ); + } ring_size =get_refill_ring_size (rq_entries); ring_ptr = mmap(NULL,
@@ -157,17 +191,23 @@ static void setup_zcrx(struct io_uring *ring) .flags = 0, }; -struct io_uring_zcrx_ifq_reg reg = { +struct t_io_uring_zcrx_ifq_reg reg = { .if_idx = ifindex, .if_rxq = cfg_queue_id, .rq_entries = rq_entries, .area_ptr = (__u64 )(unsigned long )&area_reg, .region_ptr = (__u64 )(unsigned long )®ion_reg, + .rx_buf_len = cfg_rx_buf_len, }; - ret =io_uring_register_ifq (ring, ®); -if (ret) + ret =io_uring_register_ifq (ring, (void *)®); +if (cfg_rx_buf_len && (ret == -EINVAL || ret == -EOPNOTSUPP || + ret == -ERANGE )) { +printf ("Large chunks are not supported %i\n" , ret); +exit (SKIP_CODE ); + }else if (ret) {error (1 ,0 ,"io_uring_register_ifq(): %d" , ret); + } rq_ring.khead = (unsigned int *)((char *)ring_ptr + reg.offsets.head); rq_ring.ktail = (unsigned int *)((char *)ring_ptr + reg.offsets.tail);
@@ -323,6 +363,8 @@ static void run_server(void)io_uring_queue_init (512 , &ring, flags);setup_zcrx (&ring); +if (cfg_dry_run) +return ;add_accept (&ring, fd);
@@ -383,7 +425,7 @@ static void parse_opts(int argc, char **argv)usage (argv[0 ]); cfg_payload_len = max_payload_len; -while ((c =getopt (argc, argv,"sch:p:l:i:q:o:z:" )) !=-1 ) { +while ((c =getopt (argc, argv,"sch:p:l:i:q:o:z:x:d" )) !=-1 ) {switch (c) {case 's' :if (cfg_client)
@@ -418,6 +460,12 @@ static void parse_opts(int argc, char **argv)case 'z' : cfg_send_size =strtoul (optarg,NULL ,0 );break ; +case 'x' : + cfg_rx_buf_len = page_size *strtoul (optarg,NULL ,0 ); +break ; +case 'd' : + cfg_dry_run = true; +break ; } }
tools/testing/selftests/drivers/net/hw/iou-zcrx.py +39 -0
@@ -7,6 +7,7 @@ from lib.py import ksft_run, ksft_exit, KsftSkipEx from lib.py import NetDrvEpEnv from lib.py import bkg, cmd, defer, ethtool, rand_port, wait_port_listen +SKIP_CODE = 42 def _get_current_settings(cfg): output = ethtool(f"-g {cfg.ifname}", json=True)[0]
@@ -132,6 +133,44 @@ def test_zcrx_rss(cfg) -> None: cmd(tx_cmd, host=cfg.remote) +def test_zcrx_large_chunks(cfg) -> None: + """Test zcrx with large buffer chunks.""" + + cfg.require_ipver('6') + + combined_chans = _get_combined_channels(cfg) + if combined_chans < 2: + raise KsftSkipEx('at least 2 combined channels required') + (rx_ring, hds_thresh) = _get_current_settings(cfg) + port = rand_port() + + ethtool(f"-G {cfg.ifname} tcp-data-split on") + defer(ethtool, f"-G {cfg.ifname} tcp-data-split auto") + + ethtool(f"-G {cfg.ifname} hds-thresh 0") + defer(ethtool, f"-G {cfg.ifname} hds-thresh {hds_thresh}") + + ethtool(f"-G {cfg.ifname} rx 64") + defer(ethtool, f"-G {cfg.ifname} rx {rx_ring}") + + ethtool(f"-X {cfg.ifname} equal {combined_chans - 1}") + defer(ethtool, f"-X {cfg.ifname} default") + + flow_rule_id = _set_flow_rule(cfg, port, combined_chans - 1) + defer(ethtool, f"-N {cfg.ifname} delete {flow_rule_id}") + + rx_cmd = f"{cfg.bin_local} -s -p {port} -i {cfg.ifname} -q {combined_chans - 1} -x 2" + tx_cmd = f"{cfg.bin_remote} -c -h {cfg.addr_v['6']} -p {port} -l 12840" + + probe = cmd(rx_cmd + " -d", fail=False) + if probe.ret == SKIP_CODE: + raise KsftSkipEx(probe.stdout) + + with bkg(rx_cmd, exit_wait=True): + wait_port_listen(port, proto="tcp") + cmd(tx_cmd, host=cfg.remote) + + def main() -> None: with NetDrvEpEnv(__file__) as cfg: cfg.bin_local = path.abspath(path.dirname(__file__) + "/../../../drivers/net/hw/iou-zcrx")
Signature · 1 line
2.52.0
PATCH v9 9/9 io_uring/zcrx: document area chunking parameter
In reply to Pavel Begunkovstruct io_uring_zcrx_ifq_reg::rx_buf_len is used as a hint specifying the kernel what buffer size it should use. Document the API and limitations.
Signed-off-by Pavel Begunkov <[email protected]>
--- Documentation/networking/iou-zcrx.rst | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+)
Documentation/networking/iou-zcrx.rst +20 -0
@@ -196,6 +196,26 @@ Return buffers back to the kernel to be used again:: rqe->len = cqe->res; IO_URING_WRITE_ONCE(*refill_ring.ktail, ++refill_ring.rq_tail); +Area chunking +------------- + +zcrx splits the memory area into fixed-length physically contiguous chunks. +This limits the maximum buffer size returned in a single io_uring CQE. Users +can provide a hint to the kernel to use larger chunks by setting the +``rx_buf_len`` field of ``struct io_uring_zcrx_ifq_reg`` to the desired length +during registration. If this field is set to zero, the kernel defaults to +the system page size. + +To use larger sizes, the memory area must be backed by physically contiguous +ranges whose sizes are multiples of ``rx_buf_len``. It also requires kernel +and hardware support. If registration fails, users are generally expected to +fall back to defaults by setting ``rx_buf_len`` to zero. + +Larger chunks don't give any additional guarantees about buffer sizes returned +in CQEs, and they can vary depending on many factors like traffic pattern, +hardware offload, etc. It doesn't require any application changes beyond zcrx +registration. + Testing =======
Signature · 1 line
2.52.0
Quoted 3 lines · Pavel Begunkov
Note: it's
net/only bits and doesn't include changes, which shoulf be merged separately and are posted separately. The full branch for convenience is at [1], and the patch is here:
Looks like patchwork says the patches don't apply, but the branch still merges well. Alternatively, I can rebase on top of net-next and likely delay the final io_uring commit to one release after.
Signature · 1 line
Pavel Begunkov
Jakub Kicinski
maintainerQuoted 8 lines · On Mon, 19 Jan 2026 13:54:37 +0000 Pavel Begunkov wrote:
Quoted 3 lines · Pavel Begunkov
Note: it's
net/only bits and doesn't include changes, which shoulf be merged separately and are posted separately. The full branch for convenience is at [1], and the patch is here:Looks like patchwork says the patches don't apply, but the branch still merges well. Alternatively, I can rebase on top of net-next and likely delay the final
io_uringcommit to one release after.
Yes, David refactored the test in the meantime :/
I'll pull the branch.