From: Rasesh Mody <rasesh.mody@cavium.com>
To: dev@dpdk.org
Cc: Shahed Shaikh <shahed.shaikh@cavium.com>,
ferruh.yigit@intel.com, Dept-EngDPDKDev@cavium.com
Subject: [dpdk-dev] [PATCH 15/17] net/qede: add support for Rx descriptor status
Date: Sat, 8 Sep 2018 13:31:04 -0700 [thread overview]
Message-ID: <1536438666-22184-16-git-send-email-rasesh.mody@cavium.com> (raw)
In-Reply-To: <1536438666-22184-1-git-send-email-rasesh.mody@cavium.com>
From: Shahed Shaikh <shahed.shaikh@cavium.com>
This patch implement eth_dev_ops->rx_descriptor_status
callback.
Walk through receive completion ring to calculate receive
descriptors used by firmware and then provide the status of
offset accordingly.
Signed-off-by: Shahed Shaikh <shahed.shaikh@cavium.com>
---
drivers/net/qede/qede_ethdev.c | 2 +
drivers/net/qede/qede_rxtx.c | 81 ++++++++++++++++++++++++++++++++++++++++
drivers/net/qede/qede_rxtx.h | 2 +
3 files changed, 85 insertions(+)
diff --git a/drivers/net/qede/qede_ethdev.c b/drivers/net/qede/qede_ethdev.c
index 259eb45..322400c 100644
--- a/drivers/net/qede/qede_ethdev.c
+++ b/drivers/net/qede/qede_ethdev.c
@@ -2300,6 +2300,7 @@ static int qede_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
.dev_infos_get = qede_dev_info_get,
.rx_queue_setup = qede_rx_queue_setup,
.rx_queue_release = qede_rx_queue_release,
+ .rx_descriptor_status = qede_rx_descriptor_status,
.tx_queue_setup = qede_tx_queue_setup,
.tx_queue_release = qede_tx_queue_release,
.dev_start = qede_dev_start,
@@ -2341,6 +2342,7 @@ static int qede_set_mtu(struct rte_eth_dev *dev, uint16_t mtu)
.dev_infos_get = qede_dev_info_get,
.rx_queue_setup = qede_rx_queue_setup,
.rx_queue_release = qede_rx_queue_release,
+ .rx_descriptor_status = qede_rx_descriptor_status,
.tx_queue_setup = qede_tx_queue_setup,
.tx_queue_release = qede_tx_queue_release,
.dev_start = qede_dev_start,
diff --git a/drivers/net/qede/qede_rxtx.c b/drivers/net/qede/qede_rxtx.c
index 675c0a0..8a4772f 100644
--- a/drivers/net/qede/qede_rxtx.c
+++ b/drivers/net/qede/qede_rxtx.c
@@ -2151,3 +2151,84 @@ static inline uint32_t qede_rx_cqe_to_tunn_pkt_type(uint16_t flags)
{
return 0;
}
+
+
+/* this function does a fake walk through over completion queue
+ * to calculate number of BDs used by HW.
+ * At the end, it restores the state of completion queue.
+ */
+static uint16_t
+qede_parse_fp_cqe(struct qede_rx_queue *rxq)
+{
+ uint16_t hw_comp_cons, sw_comp_cons, bd_count = 0;
+ union eth_rx_cqe *cqe, *orig_cqe = NULL;
+
+ hw_comp_cons = rte_le_to_cpu_16(*rxq->hw_cons_ptr);
+ sw_comp_cons = ecore_chain_get_cons_idx(&rxq->rx_comp_ring);
+
+ if (hw_comp_cons == sw_comp_cons)
+ return 0;
+
+ /* Get the CQE from the completion ring */
+ cqe = (union eth_rx_cqe *)ecore_chain_consume(&rxq->rx_comp_ring);
+ orig_cqe = cqe;
+
+ while (sw_comp_cons != hw_comp_cons) {
+ switch (cqe->fast_path_regular.type) {
+ case ETH_RX_CQE_TYPE_REGULAR:
+ bd_count += cqe->fast_path_regular.bd_num;
+ break;
+ case ETH_RX_CQE_TYPE_TPA_END:
+ bd_count += cqe->fast_path_tpa_end.num_of_bds;
+ break;
+ default:
+ break;
+ }
+
+ cqe =
+ (union eth_rx_cqe *)ecore_chain_consume(&rxq->rx_comp_ring);
+ sw_comp_cons = ecore_chain_get_cons_idx(&rxq->rx_comp_ring);
+ }
+
+ /* revert comp_ring to original state */
+ ecore_chain_set_cons(&rxq->rx_comp_ring, sw_comp_cons, orig_cqe);
+
+ return bd_count;
+}
+
+int
+qede_rx_descriptor_status(void *p_rxq, uint16_t offset)
+{
+ uint16_t hw_bd_cons, sw_bd_cons, sw_bd_prod;
+ uint16_t produced, consumed;
+ struct qede_rx_queue *rxq = p_rxq;
+
+ if (offset > rxq->nb_rx_desc)
+ return -EINVAL;
+
+ sw_bd_cons = ecore_chain_get_cons_idx(&rxq->rx_bd_ring);
+ sw_bd_prod = ecore_chain_get_prod_idx(&rxq->rx_bd_ring);
+
+ /* find BDs used by HW from completion queue elements */
+ hw_bd_cons = sw_bd_cons + qede_parse_fp_cqe(rxq);
+
+ if (hw_bd_cons < sw_bd_cons)
+ /* wraparound case */
+ consumed = (0xffff - sw_bd_cons) + hw_bd_cons;
+ else
+ consumed = hw_bd_cons - sw_bd_cons;
+
+ if (offset <= consumed)
+ return RTE_ETH_RX_DESC_DONE;
+
+ if (sw_bd_prod < sw_bd_cons)
+ /* wraparound case */
+ produced = (0xffff - sw_bd_cons) + sw_bd_prod;
+ else
+ produced = sw_bd_prod - sw_bd_cons;
+
+ if (offset <= produced)
+ return RTE_ETH_RX_DESC_AVAIL;
+
+ return RTE_ETH_RX_DESC_UNAVAIL;
+}
diff --git a/drivers/net/qede/qede_rxtx.h b/drivers/net/qede/qede_rxtx.h
index 8bd8d1c..d3a41e9 100644
--- a/drivers/net/qede/qede_rxtx.h
+++ b/drivers/net/qede/qede_rxtx.h
@@ -276,6 +276,8 @@ uint16_t qede_rxtx_pkts_dummy(void *p_rxq,
void qede_stop_queues(struct rte_eth_dev *eth_dev);
int qede_calc_rx_buf_size(struct rte_eth_dev *dev, uint16_t mbufsz,
uint16_t max_frame_size);
+int
+qede_rx_descriptor_status(void *rxq, uint16_t offset);
/* Fastpath resource alloc/dealloc helpers */
int qede_alloc_fp_resc(struct qede_dev *qdev);
--
1.7.10.3
next prev parent reply other threads:[~2018-09-08 20:31 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-09-08 20:30 [dpdk-dev] [PATCH 00/17] net/qede: add enhancements and fixes Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 01/17] net/qede/base: fix to handle stag update event Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 02/17] net/qede/base: add support for OneView APIs Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 03/17] net/qede/base: get pre-negotiated values for stag and bw Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 04/17] net/qede: fix to program HW regs with ether type Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 05/17] net/qede/base: limit number of non ethernet queues to 64 Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 06/17] net/qede/base: correct MCP error handler's log verbosity Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 07/17] net/qede/base: fix logic for sfp get/set Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 08/17] net/qede/base: use trust mode for forced MAC limitations Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 09/17] net/qede/base: use pointer for bytes len read Rasesh Mody
2018-09-08 20:30 ` [dpdk-dev] [PATCH 10/17] net/qede: reorganize filter code Rasesh Mody
2018-09-20 23:51 ` Ferruh Yigit
2018-09-08 20:31 ` [dpdk-dev] [PATCH 11/17] net/qede: fix flow director bug for IPv6 filter Rasesh Mody
2018-09-08 20:31 ` [dpdk-dev] [PATCH 12/17] net/qede: refactor fdir code into generic aRFS Rasesh Mody
2018-09-08 20:31 ` [dpdk-dev] [PATCH 13/17] net/qede: add support for generic flow API Rasesh Mody
2018-09-08 20:31 ` [dpdk-dev] [PATCH 14/17] net/qede: fix Rx buffer size calculation Rasesh Mody
2018-09-08 20:31 ` Rasesh Mody [this message]
2018-09-08 20:31 ` [dpdk-dev] [PATCH 16/17] net/qede/base: fix MFW FLR flow bug Rasesh Mody
2018-09-08 20:31 ` [dpdk-dev] [PATCH 17/17] net/qede: add support for dev reset Rasesh Mody
2018-09-10 5:05 ` [dpdk-dev] [PATCH 00/17] net/qede: add enhancements and fixes David Marchand
2018-09-20 16:17 ` Ferruh Yigit
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1536438666-22184-16-git-send-email-rasesh.mody@cavium.com \
--to=rasesh.mody@cavium.com \
--cc=Dept-EngDPDKDev@cavium.com \
--cc=dev@dpdk.org \
--cc=ferruh.yigit@intel.com \
--cc=shahed.shaikh@cavium.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).