[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20250227-enic_cleanup_and_ext_cq-v1-4-c314f95812bb@cisco.com>
Date: Thu, 27 Feb 2025 19:30:47 -0500
From: Satish Kharat via B4 Relay <devnull+satishkh.cisco.com@...nel.org>
To: Christian Benvenuti <benve@...co.com>,
Andrew Lunn <andrew+netdev@...n.ch>,
"David S. Miller" <davem@...emloft.net>, Eric Dumazet <edumazet@...gle.com>,
Jakub Kicinski <kuba@...nel.org>, Paolo Abeni <pabeni@...hat.com>
Cc: netdev@...r.kernel.org, linux-kernel@...r.kernel.org,
Satish Kharat <satishkh@...co.com>, Nelson Escobar <neescoba@...co.com>,
John Daley <johndale@...co.com>
Subject: [PATCH 4/8] enic: enable rq extended cq support
From: Satish Kharat <satishkh@...co.com>
Enables getting from hw all the supported rq cq sizes and
uses the highest supported cq size.
Co-developed-by: Nelson Escobar <neescoba@...co.com>
Signed-off-by: Nelson Escobar <neescoba@...co.com>
Co-developed-by: John Daley <johndale@...co.com>
Signed-off-by: John Daley <johndale@...co.com>
Signed-off-by: Satish Kharat <satishkh@...co.com>
---
drivers/net/ethernet/cisco/enic/cq_desc.h | 3 +
drivers/net/ethernet/cisco/enic/enic.h | 9 +++
drivers/net/ethernet/cisco/enic/enic_main.c | 4 ++
drivers/net/ethernet/cisco/enic/enic_res.c | 58 +++++++++++++++-
drivers/net/ethernet/cisco/enic/enic_rq.c | 101 +++++++++++++++++++++-------
5 files changed, 149 insertions(+), 26 deletions(-)
diff --git a/drivers/net/ethernet/cisco/enic/cq_desc.h b/drivers/net/ethernet/cisco/enic/cq_desc.h
index 462c5435a206b4cc93b3734fdc96a2192b53a235..8fc313b6ed0434bd55b8e10bf3086ef848acbdf1 100644
--- a/drivers/net/ethernet/cisco/enic/cq_desc.h
+++ b/drivers/net/ethernet/cisco/enic/cq_desc.h
@@ -40,6 +40,9 @@ struct cq_desc {
#define CQ_DESC_COMP_NDX_BITS 12
#define CQ_DESC_COMP_NDX_MASK ((1 << CQ_DESC_COMP_NDX_BITS) - 1)
+#define CQ_DESC_32_FI_MASK (BIT(0) | BIT(1))
+#define CQ_DESC_64_FI_MASK (BIT(0) | BIT(1))
+
static inline void cq_desc_dec(const struct cq_desc *desc_arg,
u8 *type, u8 *color, u16 *q_number, u16 *completed_index)
{
diff --git a/drivers/net/ethernet/cisco/enic/enic.h b/drivers/net/ethernet/cisco/enic/enic.h
index 305ed12aa0311ca6cc53bfbffcc300182a8011a7..d60e55accafd0e4f83728524da4f167a474d6213 100644
--- a/drivers/net/ethernet/cisco/enic/enic.h
+++ b/drivers/net/ethernet/cisco/enic/enic.h
@@ -31,6 +31,13 @@
#define ENIC_AIC_LARGE_PKT_DIFF 3
+enum ext_cq {
+ ENIC_RQ_CQ_ENTRY_SIZE_16,
+ ENIC_RQ_CQ_ENTRY_SIZE_32,
+ ENIC_RQ_CQ_ENTRY_SIZE_64,
+ ENIC_RQ_CQ_ENTRY_SIZE_MAX,
+};
+
struct enic_msix_entry {
int requested;
char devname[IFNAMSIZ + 8];
@@ -228,6 +235,7 @@ struct enic {
struct enic_rfs_flw_tbl rfs_h;
u8 rss_key[ENIC_RSS_LEN];
struct vnic_gen_stats gen_stats;
+ enum ext_cq ext_cq;
};
static inline struct net_device *vnic_get_netdev(struct vnic_dev *vdev)
@@ -349,5 +357,6 @@ int enic_is_valid_vf(struct enic *enic, int vf);
int enic_is_dynamic(struct enic *enic);
void enic_set_ethtool_ops(struct net_device *netdev);
int __enic_set_rsskey(struct enic *enic);
+void enic_ext_cq(struct enic *enic);
#endif /* _ENIC_H_ */
diff --git a/drivers/net/ethernet/cisco/enic/enic_main.c b/drivers/net/ethernet/cisco/enic/enic_main.c
index 080234ef4c2bb53c19e26601ca9bb38d26a738b7..d716514366dfc56b4e08260d18d78fddd23f6253 100644
--- a/drivers/net/ethernet/cisco/enic/enic_main.c
+++ b/drivers/net/ethernet/cisco/enic/enic_main.c
@@ -2192,6 +2192,7 @@ static void enic_reset(struct work_struct *work)
enic_init_vnic_resources(enic);
enic_set_rss_nic_cfg(enic);
enic_dev_set_ig_vlan_rewrite_mode(enic);
+ enic_ext_cq(enic);
enic_open(enic->netdev);
/* Allow infiniband to fiddle with the device again */
@@ -2218,6 +2219,7 @@ static void enic_tx_hang_reset(struct work_struct *work)
enic_init_vnic_resources(enic);
enic_set_rss_nic_cfg(enic);
enic_dev_set_ig_vlan_rewrite_mode(enic);
+ enic_ext_cq(enic);
enic_open(enic->netdev);
/* Allow infiniband to fiddle with the device again */
@@ -2592,6 +2594,8 @@ static int enic_dev_init(struct enic *enic)
enic_get_res_counts(enic);
+ enic_ext_cq(enic);
+
err = enic_alloc_enic_resources(enic);
if (err) {
dev_err(dev, "Failed to allocate enic resources\n");
diff --git a/drivers/net/ethernet/cisco/enic/enic_res.c b/drivers/net/ethernet/cisco/enic/enic_res.c
index 1261251998330c8b8363c4dd2db1ccc25847476c..a7179cc4b5296cfbce137c54a9e17e6b358a19ae 100644
--- a/drivers/net/ethernet/cisco/enic/enic_res.c
+++ b/drivers/net/ethernet/cisco/enic/enic_res.c
@@ -312,6 +312,7 @@ void enic_init_vnic_resources(struct enic *enic)
int enic_alloc_vnic_resources(struct enic *enic)
{
enum vnic_dev_intr_mode intr_mode;
+ int rq_cq_desc_size;
unsigned int i;
int err;
@@ -326,6 +327,24 @@ int enic_alloc_vnic_resources(struct enic *enic)
intr_mode == VNIC_DEV_INTR_MODE_MSIX ? "MSI-X" :
"unknown");
+ switch (enic->ext_cq) {
+ case ENIC_RQ_CQ_ENTRY_SIZE_16:
+ rq_cq_desc_size = 16;
+ break;
+ case ENIC_RQ_CQ_ENTRY_SIZE_32:
+ rq_cq_desc_size = 32;
+ break;
+ case ENIC_RQ_CQ_ENTRY_SIZE_64:
+ rq_cq_desc_size = 64;
+ break;
+ default:
+ dev_err(enic_get_dev(enic),
+ "Unable to determine rq cq desc size: %d",
+ enic->ext_cq);
+ err = -ENODEV;
+ goto err_out;
+ }
+
/* Allocate queue resources
*/
@@ -348,8 +367,8 @@ int enic_alloc_vnic_resources(struct enic *enic)
for (i = 0; i < enic->cq_count; i++) {
if (i < enic->rq_count)
err = vnic_cq_alloc(enic->vdev, &enic->cq[i], i,
- enic->config.rq_desc_count,
- sizeof(struct cq_enet_rq_desc));
+ enic->config.rq_desc_count,
+ rq_cq_desc_size);
else
err = vnic_cq_alloc(enic->vdev, &enic->cq[i], i,
enic->config.wq_desc_count,
@@ -380,6 +399,39 @@ int enic_alloc_vnic_resources(struct enic *enic)
err_out_cleanup:
enic_free_vnic_resources(enic);
-
+err_out:
return err;
}
+
+/*
+ * CMD_CQ_ENTRY_SIZE_SET can fail on older hw generations that don't support
+ * that command
+ */
+void enic_ext_cq(struct enic *enic)
+{
+ u64 a0 = CMD_CQ_ENTRY_SIZE_SET, a1 = 0;
+ int wait = 1000;
+ int ret;
+
+ spin_lock_bh(&enic->devcmd_lock);
+ ret = vnic_dev_cmd(enic->vdev, CMD_CAPABILITY, &a0, &a1, wait);
+ if (ret || a0) {
+ dev_info(&enic->pdev->dev,
+ "CMD_CQ_ENTRY_SIZE_SET not supported.");
+ enic->ext_cq = ENIC_RQ_CQ_ENTRY_SIZE_16;
+ goto out;
+ }
+ a1 &= VNIC_RQ_CQ_ENTRY_SIZE_ALL_BIT;
+ enic->ext_cq = fls(a1) - 1;
+ a0 = VNIC_RQ_ALL;
+ a1 = enic->ext_cq;
+ ret = vnic_dev_cmd(enic->vdev, CMD_CQ_ENTRY_SIZE_SET, &a0, &a1, wait);
+ if (ret) {
+ dev_info(&enic->pdev->dev, "CMD_CQ_ENTRY_SIZE_SET failed.");
+ enic->ext_cq = ENIC_RQ_CQ_ENTRY_SIZE_16;
+ }
+out:
+ spin_unlock_bh(&enic->devcmd_lock);
+ dev_info(&enic->pdev->dev, "CQ entry size set to %d bytes",
+ 16 << enic->ext_cq);
+}
diff --git a/drivers/net/ethernet/cisco/enic/enic_rq.c b/drivers/net/ethernet/cisco/enic/enic_rq.c
index 96eeb0e0d69602e8338630eab048dc960b161bbb..0013f98ccd0bfe006aea7b8a676a7872e345b8b6 100644
--- a/drivers/net/ethernet/cisco/enic/enic_rq.c
+++ b/drivers/net/ethernet/cisco/enic/enic_rq.c
@@ -21,24 +21,76 @@ static void enic_intr_update_pkt_size(struct vnic_rx_bytes_counter *pkt_size,
pkt_size->small_pkt_bytes_cnt += pkt_len;
}
-static void enic_rq_cq_desc_dec(struct cq_enet_rq_desc *desc, u8 *type,
+static void enic_rq_cq_desc_dec(void *cq_desc, u8 cq_desc_size, u8 *type,
u8 *color, u16 *q_number, u16 *completed_index)
{
/* type_color is the last field for all cq structs */
- u8 type_color = desc->type_color;
+ u8 type_color;
+
+ switch (cq_desc_size) {
+ case VNIC_RQ_CQ_ENTRY_SIZE_16: {
+ struct cq_enet_rq_desc *desc =
+ (struct cq_enet_rq_desc *)cq_desc;
+ type_color = desc->type_color;
+
+ /* Make sure color bit is read from desc *before* other fields
+ * are read from desc. Hardware guarantees color bit is last
+ * bit (byte) written. Adding the rmb() prevents the compiler
+ * and/or CPU from reordering the reads which would potentially
+ * result in reading stale values.
+ */
+ rmb();
- /* Make sure color bit is read from desc *before* other fields
- * are read from desc. Hardware guarantees color bit is last
- * bit (byte) written. Adding the rmb() prevents the compiler
- * and/or CPU from reordering the reads which would potentially
- * result in reading stale values.
- */
- rmb();
+ *q_number = le16_to_cpu(desc->q_number_rss_type_flags) &
+ CQ_DESC_Q_NUM_MASK;
+ *completed_index = le16_to_cpu(desc->completed_index_flags) &
+ CQ_DESC_COMP_NDX_MASK;
+ break;
+ }
+ case VNIC_RQ_CQ_ENTRY_SIZE_32: {
+ struct cq_enet_rq_desc_32 *desc =
+ (struct cq_enet_rq_desc_32 *)cq_desc;
+ type_color = desc->type_color;
+
+ /* Make sure color bit is read from desc *before* other fields
+ * are read from desc. Hardware guarantees color bit is last
+ * bit (byte) written. Adding the rmb() prevents the compiler
+ * and/or CPU from reordering the reads which would potentially
+ * result in reading stale values.
+ */
+ rmb();
+
+ *q_number = le16_to_cpu(desc->q_number_rss_type_flags) &
+ CQ_DESC_Q_NUM_MASK;
+ *completed_index = le16_to_cpu(desc->completed_index_flags) &
+ CQ_DESC_COMP_NDX_MASK;
+ *completed_index |= (desc->fetch_index_flags & CQ_DESC_32_FI_MASK) <<
+ CQ_DESC_COMP_NDX_BITS;
+ break;
+ }
+ case VNIC_RQ_CQ_ENTRY_SIZE_64: {
+ struct cq_enet_rq_desc_64 *desc =
+ (struct cq_enet_rq_desc_64 *)cq_desc;
+ type_color = desc->type_color;
+
+ /* Make sure color bit is read from desc *before* other fields
+ * are read from desc. Hardware guarantees color bit is last
+ * bit (byte) written. Adding the rmb() prevents the compiler
+ * and/or CPU from reordering the reads which would potentially
+ * result in reading stale values.
+ */
+ rmb();
+
+ *q_number = le16_to_cpu(desc->q_number_rss_type_flags) &
+ CQ_DESC_Q_NUM_MASK;
+ *completed_index = le16_to_cpu(desc->completed_index_flags) &
+ CQ_DESC_COMP_NDX_MASK;
+ *completed_index |= (desc->fetch_index_flags & CQ_DESC_64_FI_MASK) <<
+ CQ_DESC_COMP_NDX_BITS;
+ break;
+ }
+ }
- *q_number = le16_to_cpu(desc->q_number_rss_type_flags) &
- CQ_DESC_Q_NUM_MASK;
- *completed_index = le16_to_cpu(desc->completed_index_flags) &
- CQ_DESC_COMP_NDX_MASK;
*color = (type_color >> CQ_DESC_COLOR_SHIFT) & CQ_DESC_COLOR_MASK;
*type = type_color & CQ_DESC_TYPE_MASK;
}
@@ -113,6 +165,10 @@ static void enic_rq_set_skb_flags(struct vnic_rq *vrq, u8 type, u32 rss_hash,
}
}
+/*
+ * cq_enet_rq_desc accesses section uses only the 1st 15 bytes of the cq which is identical
+ * for all type (16,32 and 64 byte) of cqs.
+ */
static inline void cq_enet_rq_desc_dec(struct cq_enet_rq_desc *desc, u8 *ingress_port, u8 *fcoe,
u8 *eop, u8 *sop, u8 *rss_type, u8 *csum_not_calc,
u32 *rss_hash, u16 *bytes_written, u8 *packet_error,
@@ -256,7 +312,7 @@ void enic_free_rq_buf(struct vnic_rq *rq, struct vnic_rq_buf *buf)
}
static void enic_rq_indicate_buf(struct enic *enic, struct vnic_rq *rq,
- struct vnic_rq_buf *buf, struct cq_enet_rq_desc *cq_desc,
+ struct vnic_rq_buf *buf, void *cq_desc,
u8 type, u16 q_number, u16 completed_index)
{
struct sk_buff *skb;
@@ -274,7 +330,7 @@ static void enic_rq_indicate_buf(struct enic *enic, struct vnic_rq *rq,
rqstats->packets++;
- cq_enet_rq_desc_dec(cq_desc, &ingress_port,
+ cq_enet_rq_desc_dec((struct cq_enet_rq_desc *)cq_desc, &ingress_port,
&fcoe, &eop, &sop, &rss_type, &csum_not_calc,
&rss_hash, &bytes_written, &packet_error,
&vlan_stripped, &vlan_tci, &checksum, &fcoe_sof,
@@ -326,7 +382,7 @@ static void enic_rq_indicate_buf(struct enic *enic, struct vnic_rq *rq,
}
}
-static void enic_rq_service(struct enic *enic, struct cq_enet_rq_desc *cq_desc, u8 type,
+static void enic_rq_service(struct enic *enic, void *cq_desc, u8 type,
u16 q_number, u16 completed_index)
{
struct vnic_rq *vrq = &enic->rq[q_number].vrq;
@@ -358,10 +414,9 @@ unsigned int enic_rq_cq_service(struct enic *enic, unsigned int cq_index,
unsigned int work_done = 0;
struct vnic_cq *cq = &enic->cq[cq_index];
- struct cq_enet_rq_desc *cq_desc = (struct cq_enet_rq_desc *)((u8 *)cq->ring.descs
- + cq->ring.desc_size * cq->to_clean);
+ void *cq_desc = (u8 *)cq->ring.descs + cq->ring.desc_size * cq->to_clean;
- enic_rq_cq_desc_dec(cq_desc, &type, &color, &q_number,
+ enic_rq_cq_desc_dec(cq_desc, enic->ext_cq, &type, &color, &q_number,
&completed_index);
while (color != cq->last_color) {
@@ -377,10 +432,10 @@ unsigned int enic_rq_cq_service(struct enic *enic, unsigned int cq_index,
if (++work_done >= work_to_do)
break;
- cq_desc = (struct cq_enet_rq_desc *)((u8 *)cq->ring.descs
- + cq->ring.desc_size * cq->to_clean);
- enic_rq_cq_desc_dec(cq_desc, &type, &color, &q_number,
- &completed_index);
+ cq_desc = ((u8 *)cq->ring.descs +
+ cq->ring.desc_size * cq->to_clean);
+ enic_rq_cq_desc_dec(cq_desc, enic->ext_cq, &type, &color,
+ &q_number, &completed_index);
}
return work_done;
--
2.48.1
Powered by blists - more mailing lists