RDMA/mlx4: Restrict external umem for CQ when copy_to_user() is used

rleon · rleon · commit 911e5ca3e169 · 2026-03-30T13:47:45.000-04:00
When the mlx4 firmware reports the MLX4_DEV_CAP_FLAG2_SW_CQ_INIT capability, libmlx4 from the rdma-core package expects the driver to initialize memory at the address provided in the buf_addr parameter of ucmd. This behavior cannot be supported by any external umem implementation, so restrict it accordingly. Fixes: f45f195 ("RDMA/mlx4: Introduce a modern CQ creation interface") Reported-by: Jiri Pirko <jiri@nvidia.com> Link: https://patch.msgid.link/20260325-fix-mlx4-external-umem-v1-1-1c7c0e779329@nvidia.com Signed-off-by: Leon Romanovsky <leonro@nvidia.com>
diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
@@ -173,6 +173,10 @@ int mlx4_ib_create_user_cq(struct ib_cq *ibcq,
 		goto err_cq;
 	}
 
+	if (ibcq->umem &&
+	    (dev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_SW_CQ_INIT))
+		return -EOPNOTSUPP;
+
 	buf_addr = (void *)(unsigned long)ucmd.buf_addr;
 
 	if (!ibcq->umem)

Original file line number	Diff line number	Diff line change
`@@ -173,6 +173,10 @@ int mlx4_ib_create_user_cq(struct ib_cq *ibcq,`
`173`	`173`	`goto err_cq;`
`174`	`174`	`}`
`175`	`175`
	`176`	`+ if (ibcq->umem &&`
	`177`	`+ (dev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_SW_CQ_INIT))`
	`178`	`+ return -EOPNOTSUPP;`
	`179`	`+`
`176`	`180`	`buf_addr = (void *)(unsigned long)ucmd.buf_addr;`
`177`	`181`
`178`	`182`	`if (!ibcq->umem)`