i40iw: Add support to make destroy QP synchronous

Occasionally ib_write_bw crash is seen due to access of a pd object in
i40iw_sc_qp_destroy after it is freed. Destroy qp is not synchronous in
i40iw and thus the iwqp object could be referencing a pd object that is
freed by ib core as a result of successful return from i40iw_destroy_qp.

Wait in i40iw_destroy_qp till all QP references are released and destroy
the QP and its associated resources before returning.  Switch to use the
refcount API vs atomic API for lifetime management of the qp.

 RIP: 0010:i40iw_sc_qp_destroy+0x4b/0x120 [i40iw]
 [...]
 RSP: 0018:ffffb4a7042e3ba8 EFLAGS: 00010002
 RAX: 0000000000000000 RBX: 0000000000000001 RCX: dead000000000122
 RDX: ffffb4a7042e3bac RSI: ffff8b7ef9b1e940 RDI: ffff8b7efbf09080
 RBP: 0000000000000000 R08: 0000000000000001 R09: 0000000000000000
 R10: 8080808080808080 R11: 0000000000000010 R12: ffff8b7efbf08050
 R13: 0000000000000001 R14: ffff8b7f15042928 R15: ffff8b7ef9b1e940
 FS:  0000000000000000(0000) GS:ffff8b7f2fa00000(0000) knlGS:0000000000000000
 CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
 CR2: 0000000000000400 CR3: 000000020d60a006 CR4: 00000000001606e0
 Call Trace:
  i40iw_exec_cqp_cmd+0x4d3/0x5c0 [i40iw]
  ? try_to_wake_up+0x1ea/0x5d0
  ? __switch_to_asm+0x40/0x70
  i40iw_process_cqp_cmd+0x95/0xa0 [i40iw]
  i40iw_handle_cqp_op+0x42/0x1a0 [i40iw]
  ? cm_event_handler+0x13c/0x1f0 [iw_cm]
  i40iw_rem_ref+0xa0/0xf0 [i40iw]
  cm_work_handler+0x99c/0xd10 [iw_cm]
  process_one_work+0x1a1/0x360
  worker_thread+0x30/0x380
  ? process_one_work+0x360/0x360
  kthread+0x10c/0x130
  ? kthread_park+0x80/0x80
  ret_from_fork+0x35/0x40

Fixes: d37498417947 ("i40iw: add files for iwarp interface")
Link: https://lore.kernel.org/r/20200916131811.2077-1-shiraz.saleem@intel.com
Reported-by: Kamal Heib <kheib@redhat.com>
Signed-off-by: Sindhu, Devale <sindhu.devale@intel.com>
Signed-off-by: Shiraz, Saleem <shiraz.saleem@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

1 files changed, 22 insertions, 9 deletions

diff --git a/drivers/infiniband/hw/i40iw/i40iw_verbs.c b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
index 7d5ad586ce01..ffb692d619b2 100644
--- a/drivers/infiniband/hw/i40iw/i40iw_verbs.c
+++ b/drivers/infiniband/hw/i40iw/i40iw_verbs.c
@@ -364,11 +364,11 @@ static struct i40iw_pbl *i40iw_get_pbl(unsigned long va,
  * @iwqp: qp ptr (user or kernel)
  * @qp_num: qp number assigned
  */
-void i40iw_free_qp_resources(struct i40iw_device *iwdev,
-			     struct i40iw_qp *iwqp,
-			     u32 qp_num)
+void i40iw_free_qp_resources(struct i40iw_qp *iwqp)
 {
 	struct i40iw_pbl *iwpbl = &iwqp->iwpbl;
+	struct i40iw_device *iwdev = iwqp->iwdev;
+	u32 qp_num = iwqp->ibqp.qp_num;
 
 	i40iw_ieq_cleanup_qp(iwdev->vsi.ieq, &iwqp->sc_qp);
 	i40iw_dealloc_push_page(iwdev, &iwqp->sc_qp);
@@ -402,6 +402,10 @@ static void i40iw_clean_cqes(struct i40iw_qp *iwqp, struct i40iw_cq *iwcq)
 static int i40iw_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata)
 {
 	struct i40iw_qp *iwqp = to_iwqp(ibqp);
+	struct ib_qp_attr attr;
+	struct i40iw_device *iwdev = iwqp->iwdev;
+
+	memset(&attr, 0, sizeof(attr));
 
 	iwqp->destroyed = 1;
 
@@ -416,7 +420,15 @@ static int i40iw_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata)
 		}
 	}
 
-	i40iw_rem_ref(&iwqp->ibqp);
+	attr.qp_state = IB_QPS_ERR;
+	i40iw_modify_qp(&iwqp->ibqp, &attr, IB_QP_STATE, NULL);
+	i40iw_qp_rem_ref(&iwqp->ibqp);
+	wait_for_completion(&iwqp->free_qp);
+	i40iw_cqp_qp_destroy_cmd(&iwdev->sc_dev, &iwqp->sc_qp);
+	i40iw_rem_pdusecount(iwqp->iwpd, iwdev);
+	i40iw_free_qp_resources(iwqp);
+	i40iw_rem_devusecount(iwdev);
+
 	return 0;
 }
 
@@ -577,6 +589,7 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd,
 	qp->back_qp = (void *)iwqp;
 	qp->push_idx = I40IW_INVALID_PUSH_PAGE_INDEX;
 
+	iwqp->iwdev = iwdev;
 	iwqp->ctx_info.iwarp_info = &iwqp->iwarp_info;
 
 	if (i40iw_allocate_dma_mem(dev->hw,
@@ -601,7 +614,6 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd,
 		goto error;
 	}
 
-	iwqp->iwdev = iwdev;
 	iwqp->iwpd = iwpd;
 	iwqp->ibqp.qp_num = qp_num;
 	qp = &iwqp->sc_qp;
@@ -715,7 +727,7 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd,
 		goto error;
 	}
 
-	i40iw_add_ref(&iwqp->ibqp);
+	refcount_set(&iwqp->refcount, 1);
 	spin_lock_init(&iwqp->lock);
 	iwqp->sig_all = (init_attr->sq_sig_type == IB_SIGNAL_ALL_WR) ? 1 : 0;
 	iwdev->qp_table[qp_num] = iwqp;
@@ -737,10 +749,11 @@ static struct ib_qp *i40iw_create_qp(struct ib_pd *ibpd,
 	}
 	init_completion(&iwqp->sq_drained);
 	init_completion(&iwqp->rq_drained);
+	init_completion(&iwqp->free_qp);
 
 	return &iwqp->ibqp;
 error:
-	i40iw_free_qp_resources(iwdev, iwqp, qp_num);
+	i40iw_free_qp_resources(iwqp);
 	return ERR_PTR(err_code);
 }
 
@@ -2629,13 +2642,13 @@ static const struct ib_device_ops i40iw_dev_ops = {
 	.get_hw_stats = i40iw_get_hw_stats,
 	.get_port_immutable = i40iw_port_immutable,
 	.iw_accept = i40iw_accept,
-	.iw_add_ref = i40iw_add_ref,
+	.iw_add_ref = i40iw_qp_add_ref,
 	.iw_connect = i40iw_connect,
 	.iw_create_listen = i40iw_create_listen,
 	.iw_destroy_listen = i40iw_destroy_listen,
 	.iw_get_qp = i40iw_get_qp,
 	.iw_reject = i40iw_reject,
-	.iw_rem_ref = i40iw_rem_ref,
+	.iw_rem_ref = i40iw_qp_rem_ref,
 	.map_mr_sg = i40iw_map_mr_sg,
 	.mmap = i40iw_mmap,
 	.modify_qp = i40iw_modify_qp,