diff mbox series

[net-next,08/11] bnxt_en: Implement RX ring reset in response to buffer errors.

Message ID 1601839381-10446-9-git-send-email-michael.chan@broadcom.com
State New
Headers show
Series bnxt_en: net-next updates. | expand

Commit Message

Michael Chan Oct. 4, 2020, 7:22 p.m. UTC
On some older chips, it is necessary to do a reset when we get buffer
errors associated with an RX ring.  These buffer errors may become
frequent if the RX ring underruns under heavy traffic.  The current
code does a global reset of all reasources when this happens.  This
works but creates a big disruption of all rings when one RX ring is
having problem.  This patch implements a localized RX ring reset of
just the RX ring having the issue.  All other rings including all
TX rings will not be affected by this single RX ring reset.

Only the older chips prior to the P5 class supports this reset.
Because it is not a global reset, packets may still be arriving
while we are calling firmware to reset that ring.  We need to be
sure that we don't post any buffers during this time while the
ring is undergoing reset.  After firmware completes successfully,
the ring will be in the reset state with no buffers and we can start
filling it with new buffers and posting them.

Reviewed-by: Pavan Chebbi <pavan.chebbi@broadcom.com>
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.c | 79 +++++++++++++++++++++--
 1 file changed, 75 insertions(+), 4 deletions(-)
diff mbox series

Patch

diff --git a/drivers/net/ethernet/broadcom/bnxt/bnxt.c b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
index ef0267060a46..31877fb4ddd9 100644
--- a/drivers/net/ethernet/broadcom/bnxt/bnxt.c
+++ b/drivers/net/ethernet/broadcom/bnxt/bnxt.c
@@ -1172,7 +1172,10 @@  static void bnxt_sched_reset(struct bnxt *bp, struct bnxt_rx_ring_info *rxr)
 {
 	if (!rxr->bnapi->in_reset) {
 		rxr->bnapi->in_reset = true;
-		set_bit(BNXT_RESET_TASK_SP_EVENT, &bp->sp_event);
+		if (bp->flags & BNXT_FLAG_CHIP_P5)
+			set_bit(BNXT_RESET_TASK_SP_EVENT, &bp->sp_event);
+		else
+			set_bit(BNXT_RST_RING_SP_EVENT, &bp->sp_event);
 		bnxt_queue_sp_work(bp);
 	}
 	rxr->rx_next_cons = 0xffff;
@@ -1773,8 +1776,8 @@  static int bnxt_rx_pkt(struct bnxt *bp, struct bnxt_cp_ring_info *cpr,
 		if (rx_err & RX_CMPL_ERRORS_BUFFER_ERROR_MASK) {
 			bnapi->cp_ring.sw_stats.rx.rx_buf_errors++;
 			if (!(bp->flags & BNXT_FLAG_CHIP_P5)) {
-				netdev_warn(bp->dev, "RX buffer error %x\n",
-					    rx_err);
+				netdev_warn_once(bp->dev, "RX buffer error %x\n",
+						 rx_err);
 				bnxt_sched_reset(bp, rxr);
 			}
 		}
@@ -2250,7 +2253,7 @@  static void __bnxt_poll_work_done(struct bnxt *bp, struct bnxt_napi *bnapi)
 		bnapi->tx_pkts = 0;
 	}
 
-	if (bnapi->events & BNXT_RX_EVENT) {
+	if ((bnapi->events & BNXT_RX_EVENT) && !(bnapi->in_reset)) {
 		struct bnxt_rx_ring_info *rxr = bnapi->rx_ring;
 
 		if (bnapi->events & BNXT_AGG_EVENT)
@@ -10510,6 +10513,23 @@  static void bnxt_dbg_dump_states(struct bnxt *bp)
 	}
 }
 
+static int bnxt_hwrm_rx_ring_reset(struct bnxt *bp, int ring_nr)
+{
+	struct bnxt_rx_ring_info *rxr = &bp->rx_ring[ring_nr];
+	struct hwrm_ring_reset_input req = {0};
+	struct bnxt_napi *bnapi = rxr->bnapi;
+	struct bnxt_cp_ring_info *cpr;
+	u16 cp_ring_id;
+
+	cpr = &bnapi->cp_ring;
+	cp_ring_id = cpr->cp_ring_struct.fw_ring_id;
+	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_RING_RESET, cp_ring_id, -1);
+	req.ring_type = RING_RESET_REQ_RING_TYPE_RX_RING_GRP;
+	req.ring_id = cpu_to_le16(bp->grp_info[bnapi->index].fw_grp_id);
+	return hwrm_send_message_silent(bp, &req, sizeof(req),
+					HWRM_CMD_TIMEOUT);
+}
+
 static void bnxt_reset_task(struct bnxt *bp, bool silent)
 {
 	if (!silent)
@@ -10645,6 +10665,54 @@  static void bnxt_reset(struct bnxt *bp, bool silent)
 	bnxt_rtnl_unlock_sp(bp);
 }
 
+/* Only called from bnxt_sp_task() */
+static void bnxt_rx_ring_reset(struct bnxt *bp)
+{
+	int i;
+
+	bnxt_rtnl_lock_sp(bp);
+	if (!test_bit(BNXT_STATE_OPEN, &bp->state)) {
+		bnxt_rtnl_unlock_sp(bp);
+		return;
+	}
+	/* Disable and flush TPA before resetting the RX ring */
+	if (bp->flags & BNXT_FLAG_TPA)
+		bnxt_set_tpa(bp, false);
+	for (i = 0; i < bp->rx_nr_rings; i++) {
+		struct bnxt_rx_ring_info *rxr = &bp->rx_ring[i];
+		struct bnxt_cp_ring_info *cpr;
+		int rc;
+
+		if (!rxr->bnapi->in_reset)
+			continue;
+
+		rc = bnxt_hwrm_rx_ring_reset(bp, i);
+		if (rc) {
+			if (rc == -EINVAL || rc == -EOPNOTSUPP)
+				netdev_info_once(bp->dev, "RX ring reset not supported by firmware, falling back to global reset\n");
+			else
+				netdev_warn(bp->dev, "RX ring reset failed, rc = %d, falling back to global reset\n",
+					    rc);
+			bnxt_reset_task(bp, false);
+			break;
+		}
+		bnxt_free_one_rx_ring_skbs(bp, i);
+		rxr->rx_prod = 0;
+		rxr->rx_agg_prod = 0;
+		rxr->rx_sw_agg_prod = 0;
+		rxr->rx_next_cons = 0;
+		rxr->bnapi->in_reset = false;
+		bnxt_alloc_one_rx_ring(bp, i);
+		cpr = &rxr->bnapi->cp_ring;
+		if (bp->flags & BNXT_FLAG_AGG_RINGS)
+			bnxt_db_write(bp, &rxr->rx_agg_db, rxr->rx_agg_prod);
+		bnxt_db_write(bp, &rxr->rx_db, rxr->rx_prod);
+	}
+	if (bp->flags & BNXT_FLAG_TPA)
+		bnxt_set_tpa(bp, true);
+	bnxt_rtnl_unlock_sp(bp);
+}
+
 static void bnxt_fw_reset_close(struct bnxt *bp)
 {
 	bnxt_ulp_stop(bp);
@@ -10933,6 +11001,9 @@  static void bnxt_sp_task(struct work_struct *work)
 	if (test_and_clear_bit(BNXT_RESET_TASK_SILENT_SP_EVENT, &bp->sp_event))
 		bnxt_reset(bp, true);
 
+	if (test_and_clear_bit(BNXT_RST_RING_SP_EVENT, &bp->sp_event))
+		bnxt_rx_ring_reset(bp);
+
 	if (test_and_clear_bit(BNXT_FW_RESET_NOTIFY_SP_EVENT, &bp->sp_event))
 		bnxt_devlink_health_report(bp, BNXT_FW_RESET_NOTIFY_SP_EVENT);