STRUCT IDPF_Q_VEC_RSRC(9)

💖 struct idpf_q_vec_rsrc(9) 💖

okay so you're reading the Kernel API Manual (that's the manual name bestie, we're in kernel-doc land now) and this page is literally all about struct idpf_q_vec_rsrc — the handle for queue and vector resources. it's giving "one struct to hold every queue-and-vector thing you'll ever need" and honestly? iconic. 🎀

---

✨ NAME ✨

struct idpf_q_vec_rsrc — handle for queue and vector resources 🦋

---

✨ SYNOPSIS ✨

struct idpf_q_vec_rsrc {
    struct device           *dev;
    struct idpf_q_vector    *q_vectors;
    u16 *q_vector_idxs;
    u16 num_q_vectors;
    u16 noirq_v_idx;
    u32 noirq_dyn_ctl_ena;
    void __iomem            *noirq_dyn_ctl;
    struct idpf_txq_group   *txq_grps;
    u32 txq_desc_count;
    u32 complq_desc_count;
    u32 txq_model;
    u16 num_txq;
    u16 num_complq;
    u16 num_txq_grp;
    u16 xdp_txq_offset;
    u16 num_rxq_grp;
    u32 rxq_model;
    struct idpf_rxq_group   *rxq_grps;
    u16 num_rxq;
    u16 num_bufq;
    u32 rxq_desc_count;
    u32 bufq_desc_count[IDPF_MAX_BUFQS_PER_RXQ_GRP];
    u8 num_bufqs_per_qgrp;
    bool base_rxd;
 };

look at that layout. absolutely ate. no cap, every single field is right there and we're gonna go through all of them together 💅

---

🌸 Members 🌸

🧚 The device + vector side

• dev — device pointer for DMA mapping. like, the actual thing you're mapping DMA against 🫶
• q_vectors — array of queue vectors. the whole squad of them.
• q_vector_idxs — starting index of queue vectors.
• num_q_vectors — number of IRQ vectors allocated. count 'em.
• noirq_v_idx — ID of the NOIRQ vector.
• noirq_dyn_ctl_ena — the value to write to the above to enable it. (the above = that register right below, keep reading lol)
• noirq_dyn_ctl — register to enable/disable the vector for NOIRQ queues 🔥

🎀 The TX side

• txq_grps — array of TX queue groups.
• txq_desc_count — TX queue descriptor count.
• complq_desc_count — completion queue descriptor count.
• txq_model — split queue or single queue queuing model.
• num_txq — number of allocated TX queues.
• num_complq — number of allocated completion queues.
• num_txq_grp — number of TX queue groups.
• xdp_txq_offset — index of the first XDPSQ (== number of regular SQs). yes the double equals is literally in the original, we respect it 💜

🦋 The RX side

• num_rxq_grp — number of RX queues in a group.
• rxq_model — splitq queue or single queue queuing model. (their spelling, not mine, keeping it 💀)
• rxq_grps — total number of RX groups. Number of groups * number of RX per group will yield total number of RX queues. math slay 🧮✨
• num_rxq — number of allocated RX queues.
• num_bufq — number of allocated buffer queues.
• rxq_desc_count — RX queue descriptor count. *MUST* have enough descriptors to complete all buffer descriptors for all buffer queues in the worst case. like, genuinely must. that's an all-caps MUST, do not play with it 🙈
• bufq_desc_count — buffer queue descriptor count. it's an array sized IDPF_MAX_BUFQS_PER_RXQ_GRP, just saying 👀
• num_bufqs_per_qgrp — buffer queues per RX queue in a given grouping.
• base_rxd — true if the driver should use base descriptors instead of flex.

fr fr the only scary one here is rxq_desc_count — that MUST is doing a lot of heavy lifting. give it enough descriptors or the worst case will find you. 😭

---

🔥 SEE ALSO 🔥

Kernel file ./drivers/net/ethernet/intel/idpf/idpf.h — and bestie, the whole friend group is listed, we're going through all of them:

• struct idpf_mac_filter(9)
• enum idpf_state(9)
• enum idpf_flags(9)
• enum idpf_cap_field(9)
• enum idpf_vport_state(9)
• struct idpf_netdev_priv(9)
• struct idpf_reset_reg(9)
• struct idpf_vport_max_q(9)
• struct idpf_reg_ops(9)
• struct idpf_dev_ops(9)
• enum idpf_vport_reset_cause(9)
• enum idpf_vport_flags(9)
• struct idpf_tstamp_stats(9)
• struct idpf_vport(9)
• enum idpf_user_flags(9)
• struct idpf_rss_data(9)
• struct idpf_q_coalesce(9)
• struct idpf_vport_user_config_data(9)
• enum idpf_vport_config_flags(9)
• struct idpf_avail_queue_info(9)
• struct idpf_vector_info(9)
• struct idpf_vector_lifo(9)
• struct idpf_queue_id_reg_chunk(9)
• struct idpf_queue_id_reg_info(9)
• struct idpf_vport_config(9)
• struct idpf_adapter(9)
• idpf_is_queue_model_split(9)
• idpf_is_rdma_cap_ena(9)
• idpf_get_reserved_vecs(9)
• idpf_get_reserved_rdma_vecs(9)
• idpf_get_default_vports(9)
• idpf_get_max_vports(9)
• idpf_get_max_tx_bufs(9)
• idpf_get_min_tx_pkt_len(9)
• idpf_is_reset_detected(9)
• idpf_is_reset_in_prog(9)
• idpf_netdev_to_vport(9)
• idpf_netdev_to_adapter(9)
• idpf_is_feature_ena(9)
• idpf_get_max_tx_hdr_size(9)
• idpf_vport_ctrl_lock(9)
• idpf_vport_ctrl_unlock(9)

like?? that's a whole group chat of references and every single one of them made the list. obsessed. 👑

---

and that's the page!! source is ./drivers/net/ethernet/intel/idpf, dated October 2026, and it's living in the Kernel API Manual — mwah, bye bestie, go read your queue descriptors 💜✨🎀