Unverified Commit 5de0187d authored by openeuler-ci-bot's avatar openeuler-ci-bot Committed by Gitee
Browse files

!8396 RDMA/hns: A set of bugfixes for openeuler 22.03 SP4 

Merge Pull Request from: @ygf_123 
 
A set of bugfixes for openeuler 22.03 SP4

Chengchang Tang (4):
  RDMA/hns: Fix cpu stuck by printings during reset
  RDMA/hns: Fix missing default values of scc parameters
  RDMA/hns: directly read the scc parameters in the driver instead of
    the firmware
  RDMA/hns: Fix deadlock when using DCA debugfs.

Junxian Huang (2):
  RDMA/hns: Fix missing validation check of DCA mem size
  RDMA/hns: Set NULL after mtr_node is freed

wenglianfa (3):
  RDMA/hns: Fix CPU stuck due to read polling during reset
  RDMA/hns: Fix fmt output error
  RDMA/hns: Fix missing list_del in hns_roce_hem

https://gitee.com/openeuler/kernel/issues/I9SI44 
 
Link:https://gitee.com/openeuler/kernel/pulls/8396

 

Reviewed-by: default avatarChengchang Tang <tangchengchang@huawei.com>
Signed-off-by: default avatarJialin Zhang <zhangjialin11@huawei.com>
parents 16e09ea7 f5e0033c

drivers/infiniband/hw/hns/hns_roce_cq.c

+5 −2
Original line number Diff line number Diff line
@@ -179,8 +179,9 @@ static void free_cqc(struct hns_roce_dev *hr_dev, struct hns_roce_cq *hr_cq) 
	ret = hns_roce_destroy_hw_ctx(hr_dev, HNS_ROCE_CMD_DESTROY_CQC,

				      hr_cq->cqn);

	if (ret)

		dev_err(dev, "DESTROY_CQ failed (%d) for CQN %06lx\n", ret,

			hr_cq->cqn);

		dev_err_ratelimited(dev,

				    "DESTROY_CQ failed (%d) for CQN %06lx\n",

				    ret, hr_cq->cqn);

	if (ret == -EBUSY)

		hr_cq->delayed_destroy_flag = true;
@@ -223,6 +224,7 @@ static int alloc_cq_buf(struct hns_roce_dev *hr_dev, struct hns_roce_cq *hr_cq, 
	if (ret) {

		ibdev_err(ibdev, "Failed to alloc CQ mtr, ret = %d\n", ret);

		kvfree(hr_cq->mtr_node);

		hr_cq->mtr_node = NULL;

	}



	return ret;
@@ -235,6 +237,7 @@ static void free_cq_buf(struct hns_roce_dev *hr_dev, struct hns_roce_cq *hr_cq) 
	} else {

		hns_roce_mtr_destroy(hr_dev, &hr_cq->mtr);

		kvfree(hr_cq->mtr_node);

		hr_cq->mtr_node = NULL;

	}

}

drivers/infiniband/hw/hns/hns_roce_dca.c

+5 −0
Original line number Diff line number Diff line
@@ -1529,6 +1529,11 @@ static int UVERBS_HANDLER(HNS_IB_METHOD_DCA_MEM_REG)( 
	if (ret)

		return ret;



	if (!init_attr.size)

		return -EINVAL;



	init_attr.size = hr_hw_page_align(init_attr.size);



	mem = alloc_dca_mem(to_hr_dca_ctx(hr_dev, uctx));

	if (!mem)

		return -ENOMEM;

drivers/infiniband/hw/hns/hns_roce_debugfs.c

+6 −6
Original line number Diff line number Diff line
@@ -353,19 +353,19 @@ static void dca_stats_dev_qp_in_seqfile(struct hns_roce_dev *hr_dev, 
	seq_printf(file, "%-10s %-10s %-10s %s\n", "QPN", "Size(kB)", "PID",

		   "State");



	xa_lock(&hr_dev->qp_table_xa);

	xa_lock_irq(&hr_dev->qp_table_xa);

	xa_for_each(&hr_dev->qp_table_xa, id, hr_qp) {

		stats.total_size = 0;

		dca_setup_qp_stats(hr_qp, &stats);

		if (!stats.total_size)

			continue;



		xa_unlock(&hr_dev->qp_table_xa);

		xa_unlock_irq(&hr_dev->qp_table_xa);

		seq_printf(file, "%-10u %-10u %-10s %-s\n", stats.qpn,

			   stats.total_size / KB, stats.name, stats.state);

		xa_lock(&hr_dev->qp_table_xa);

		xa_lock_irq(&hr_dev->qp_table_xa);

	}

	xa_unlock(&hr_dev->qp_table_xa);

	xa_unlock_irq(&hr_dev->qp_table_xa);

}



static void dca_stats_ctx_qp_in_seqfile(struct hns_roce_dev *hr_dev,
@@ -391,11 +391,11 @@ static void dca_stats_ctx_qp_in_seqfile(struct hns_roce_dev *hr_dev, 
	dca_ctx_stats_qp(ctx, bitmap, nbits);

	for_each_set_bit(qpn, bitmap, nbits) {

		stats.total_size = 0;

		xa_lock(&hr_dev->qp_table_xa);

		xa_lock_irq(&hr_dev->qp_table_xa);

		hr_qp = __hns_roce_qp_lookup(hr_dev, qpn);

		if (hr_qp)

			dca_setup_qp_stats(hr_qp, &stats);

		xa_unlock(&hr_dev->qp_table_xa);

		xa_unlock_irq(&hr_dev->qp_table_xa);

		if (!stats.total_size)

			continue;

drivers/infiniband/hw/hns/hns_roce_hem.c

+4 −2
Original line number Diff line number Diff line
@@ -321,6 +321,7 @@ void hns_roce_free_hem(struct hns_roce_dev *hr_dev, struct hns_roce_hem *hem) 
		return;



	list_for_each_entry_safe(chunk, tmp, &hem->chunk_list, list) {

		list_del(&chunk->list);

		for (i = 0; i < chunk->npages; ++i)

			dma_free_coherent(hr_dev->dev,

				   sg_dma_len(&chunk->mem[i]),
@@ -712,7 +713,8 @@ void hns_roce_table_put(struct hns_roce_dev *hr_dev, 


	ret = hr_dev->hw->clear_hem(hr_dev, table, obj, HEM_HOP_STEP_DIRECT);

	if (ret)

		dev_warn(dev, "failed to clear HEM base address, ret = %d.\n",

		dev_warn_ratelimited(dev,

			"failed to clear HEM base address, ret = %d.\n",

			ret);



	hns_roce_free_hem(hr_dev, table->hem[i]);

drivers/infiniband/hw/hns/hns_roce_hw_v2.c

+34 −32
Original line number Diff line number Diff line
@@ -33,7 +33,6 @@ 
#include <linux/acpi.h>

#include <linux/etherdevice.h>

#include <linux/interrupt.h>

#include <linux/iopoll.h>

#include <linux/kernel.h>

#include <linux/types.h>

#include <net/addrconf.h>
@@ -452,11 +451,13 @@ static int check_send_valid(struct hns_roce_dev *hr_dev, 
	if (unlikely(hr_qp->state == IB_QPS_RESET ||

		     hr_qp->state == IB_QPS_INIT ||

		     hr_qp->state == IB_QPS_RTR)) {

		ibdev_err(ibdev, "failed to post WQE, QP state %u!\n",

		ibdev_err_ratelimited(ibdev,

				      "failed to post WQE, QP state %u!\n",

				      hr_qp->state);

		return -EINVAL;

	} else if (unlikely(hr_dev->state >= HNS_ROCE_DEVICE_STATE_RST_DOWN)) {

		ibdev_err(ibdev, "failed to post WQE, dev state %d!\n",

		ibdev_err_ratelimited(ibdev,

				      "failed to post WQE, dev state %d!\n",

				      hr_dev->state);

		return -EIO;

	}
@@ -464,8 +465,8 @@ static int check_send_valid(struct hns_roce_dev *hr_dev, 
	if (check_dca_attach_enable(hr_qp)) {

		ret = dca_attach_qp_buf(hr_dev, hr_qp);

		if (unlikely(ret)) {

			ibdev_err(ibdev,

				  "failed to attach DCA for QP-%ld send!\n",

			ibdev_err_ratelimited(ibdev,

				  "failed to attach DCA for QP-%lu send!\n",

				  hr_qp->qpn);

			return ret;

		}
@@ -1146,14 +1147,9 @@ static u32 hns_roce_v2_cmd_hw_resetting(struct hns_roce_dev *hr_dev, 
					unsigned long instance_stage,

					unsigned long reset_stage)

{

#define HW_RESET_TIMEOUT_US 1000000

#define HW_RESET_DELAY_US 1



	struct hns_roce_v2_priv *priv = hr_dev->priv;

	struct hnae3_handle *handle = priv->handle;

	const struct hnae3_ae_ops *ops = handle->ae_algo->ops;

	unsigned long val;

	int ret;



	/* When hardware reset is detected, we should stop sending mailbox&cmq&

	 * doorbell to hardware. If now in .init_instance() function, we should
@@ -1166,10 +1162,7 @@ static u32 hns_roce_v2_cmd_hw_resetting(struct hns_roce_dev *hr_dev, 
	 */

	hr_dev->dis_db = true;



	ret = read_poll_timeout_atomic(ops->ae_dev_reset_cnt, val,

				val > hr_dev->reset_cnt, HW_RESET_DELAY_US,

				HW_RESET_TIMEOUT_US, false, handle);

	if (!ret)

	if (!ops->get_hw_reset_stat(handle))

		hr_dev->is_reset = true;



	if (!hr_dev->is_reset || reset_stage == HNS_ROCE_STATE_RST_INIT ||
@@ -3827,7 +3820,8 @@ static int free_mr_post_send_lp_wqe(struct hns_roce_qp *hr_qp) 


	ret = hns_roce_v2_post_send(&hr_qp->ibqp, send_wr, &bad_wr);

	if (ret) {

		ibdev_err(ibdev, "failed to post wqe for free mr, ret = %d.\n",

		ibdev_err_ratelimited(ibdev,

			"failed to post wqe for free mr, ret = %d.\n",

			ret);

		return ret;

	}
@@ -3867,7 +3861,7 @@ static void free_mr_send_cmd_to_hw(struct hns_roce_dev *hr_dev) 


		ret = free_mr_post_send_lp_wqe(hr_qp);

		if (ret) {

			ibdev_err(ibdev,

			ibdev_err_ratelimited(ibdev,

				  "failed to send wqe (qp:0x%lx) for free mr, ret = %d.\n",

				  hr_qp->qpn, ret);

			break;
@@ -3880,14 +3874,14 @@ static void free_mr_send_cmd_to_hw(struct hns_roce_dev *hr_dev) 
	while (cqe_cnt) {

		npolled = hns_roce_v2_poll_cq(&free_mr->rsv_cq->ib_cq, cqe_cnt, wc);

		if (npolled < 0) {

			ibdev_err(ibdev,

			ibdev_err_ratelimited(ibdev,

				  "failed to poll cqe for free mr, remain %d cqe.\n",

				  cqe_cnt);

			goto out;

		}



		if (time_after(jiffies, end)) {

			ibdev_err(ibdev,

			ibdev_err_ratelimited(ibdev,

				  "failed to poll cqe for free mr and timeout, remain %d cqe.\n",

				  cqe_cnt);

			goto out;
@@ -5550,7 +5544,8 @@ static int hns_roce_v2_set_abs_fields(struct ib_qp *ibqp, 
	int ret = 0;



	if (!check_qp_state(cur_state, new_state)) {

		ibdev_err(&hr_dev->ib_dev, "Illegal state for QP!\n");

		ibdev_err_ratelimited(&hr_dev->ib_dev,

				      "Illegal state for QP!\n");

		return -EINVAL;

	}
@@ -5810,7 +5805,7 @@ static int hns_roce_v2_modify_qp(struct ib_qp *ibqp, 
	/* SW pass context to HW */

	ret = hns_roce_v2_qp_modify(hr_dev, context, qpc_mask, hr_qp);

	if (ret) {

		ibdev_err(ibdev, "failed to modify QP, ret = %d.\n", ret);

		ibdev_err_ratelimited(ibdev, "failed to modify QP, ret = %d.\n", ret);

		goto out;

	}
@@ -5995,7 +5990,9 @@ static int hns_roce_v2_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, 


	ret = hns_roce_v2_query_qpc(hr_dev, hr_qp->qpn, &context);

	if (ret) {

		ibdev_err(ibdev, "failed to query QPC, ret = %d.\n", ret);

		ibdev_err_ratelimited(ibdev,

				      "failed to query QPC, ret = %d.\n",

				      ret);

		ret = -EINVAL;

		goto out;

	}
@@ -6003,7 +6000,7 @@ static int hns_roce_v2_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, 
	state = hr_reg_read(&context, QPC_QP_ST);

	tmp_qp_state = to_ib_qp_st((enum hns_roce_v2_qp_state)state);

	if (tmp_qp_state == -1) {

		ibdev_err(ibdev, "Illegal ib_qp_state\n");

		ibdev_err_ratelimited(ibdev, "Illegal ib_qp_state\n");

		ret = -EINVAL;

		goto out;

	}
@@ -6085,7 +6082,9 @@ static bool hns_roce_v2_chk_dca_buf_inactive(struct hns_roce_dev *hr_dev, 


	ret = hns_roce_v2_query_qpc(hr_dev, hr_qp->qpn, &context);

	if (ret) {

		ibdev_err(ibdev, "failed to query DCA QPC, ret = %d.\n", ret);

		ibdev_err_ratelimited(ibdev,

				      "failed to query DCA QPC, ret = %d.\n",

				      ret);

		return false;

	}
@@ -6129,7 +6128,7 @@ int hns_roce_v2_destroy_qp_common(struct hns_roce_dev *hr_dev, 
		ret = hns_roce_v2_modify_qp(&hr_qp->ibqp, NULL, 0,

					    hr_qp->state, IB_QPS_RESET, udata);

		if (ret)

			ibdev_err(ibdev,

			ibdev_err_ratelimited(ibdev,

				  "failed to modify QP to RST, ret = %d.\n",

				  ret);

	}
@@ -6167,7 +6166,7 @@ int hns_roce_v2_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata) 


	ret = hns_roce_v2_destroy_qp_common(hr_dev, hr_qp, udata);

	if (ret)

		ibdev_err(&hr_dev->ib_dev,

		ibdev_err_ratelimited(&hr_dev->ib_dev,

			  "failed to destroy QP, QPN = 0x%06lx, ret = %d.\n",

			  hr_qp->qpn, ret);
@@ -6466,7 +6465,7 @@ static int hns_roce_v2_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period) 
				HNS_ROCE_CMD_MODIFY_CQC, hr_cq->cqn);

	hns_roce_free_cmd_mailbox(hr_dev, mailbox);

	if (ret)

		ibdev_err(&hr_dev->ib_dev,

		ibdev_err_ratelimited(&hr_dev->ib_dev,

			  "failed to process cmd when modifying CQ, ret = %d.\n",

			  ret);
@@ -6492,7 +6491,7 @@ static int hns_roce_v2_query_cqc(struct hns_roce_dev *hr_dev, u32 cqn, 
	ret = hns_roce_cmd_mbox(hr_dev, 0, mailbox->dma,

				HNS_ROCE_CMD_QUERY_CQC, cqn);

	if (ret) {

		ibdev_err(&hr_dev->ib_dev,

		ibdev_err_ratelimited(&hr_dev->ib_dev,

			"failed to process cmd when querying CQ, ret = %d.\n",

			ret);

		goto err_mailbox;
@@ -7389,6 +7388,9 @@ static int hns_roce_v2_query_scc_param(struct hns_roce_dev *hr_dev, 
	struct hns_roce_port *pdata;

	int ret;



	if (hr_dev->pci_dev->revision <= PCI_REVISION_ID_HIP08 || hr_dev->is_vf)

		return -EOPNOTSUPP;



	if (port_num > hr_dev->caps.num_ports) {

		ibdev_err_ratelimited(&hr_dev->ib_dev,

				      "invalid port num %u.\n", port_num);