RDMA/iw_cxgb4: Fix refcount underflow while destroying cqs.

Previous atomic increment/decrement logic expects the atomic count to be
'0' after the final decrement.

Replacing atomic count with refcount does not allow that, as
refcount_dec() considers count of 1 as underflow and triggers a kernel
splat.

Fix the current refcount logic by using the usual pattern of decrementing
the refcount and test if it is '0' on the final deref in
c4iw_destroy_cq(). Use wait_for_completion() instead of wait_event().

Fixes: 7183451f846d ("RDMA/cxgb4: Use refcount_t instead of atomic_t for reference counting")
Link: https://lore.kernel.org/r/1628167412-12114-1-git-send-email-dakshaja@chelsio.com
Signed-off-by: Dakshaja Uppalapati <dakshaja@chelsio.com>
Reviewed-by: Potnuri Bharat Teja <bharat@chelsio.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

1 files changed, 2 insertions, 1 deletions

diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
index 3883af3d2312..ac5f581aff4c 100644
--- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
+++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
@@ -428,7 +428,7 @@ struct c4iw_cq {
 	spinlock_t lock;
 	spinlock_t comp_handler_lock;
 	refcount_t refcnt;
-	wait_queue_head_t wait;
+	struct completion cq_rel_comp;
 	struct c4iw_wr_wait *wr_waitp;
 };
 
@@ -979,6 +979,7 @@ struct ib_mr *c4iw_reg_user_mr(struct ib_pd *pd, u64 start,
 struct ib_mr *c4iw_get_dma_mr(struct ib_pd *pd, int acc);
 int c4iw_dereg_mr(struct ib_mr *ib_mr, struct ib_udata *udata);
 int c4iw_destroy_cq(struct ib_cq *ib_cq, struct ib_udata *udata);
+void c4iw_cq_rem_ref(struct c4iw_cq *chp);
 int c4iw_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr,
 		   struct ib_udata *udata);
 int c4iw_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags);