Re: [PATCH] Revert "scsi: iscsi: Fix HW conn removal use after free"
5 views
Skip to first unread message
Mike Christie
Jul 17, 2025, 4:01:33 PMJul 17
to Li Lingfeng, ldu...@suse.com, cle...@redhat.com, nja...@marvell.com, mrang...@marvell.com, GR-QLogic-Sto...@marvell.com, martin....@oracle.com, je...@linux.ibm.com, James.B...@hansenpartnership.com, open-...@googlegroups.com, linux...@vger.kernel.org, linux-...@vger.kernel.org, yuk...@huaweicloud.com, hou...@huawei.com, yi.z...@huawei.com, yang...@huawei.com, chengz...@huawei.com, lilin...@huaweicloud.com
On 7/15/25 2:39 AM, Li Lingfeng wrote:
> This reverts commit c577ab7ba5f3bf9062db8a58b6e89d4fe370447e.
>
> The invocation of iscsi_put_conn() in iscsi_iter_destory_conn_fn() is used
> to free the initial reference counter of iscsi_cls_conn.
> For non-qla4xxx cases, the ->destroy_conn() callback (e.g.,
> iscsi_conn_teardown) will call iscsi_remove_conn() and iscsi_put_conn() to
> remove the connection from the children list of session and free the
> connection at last.
> However for qla4xxx, it is not the case. The ->destroy_conn() callback
> of qla4xxx will keep the connection in the session conn_list and doesn't
> use iscsi_put_conn() to free the initial reference counter. Therefore,
> it seems necessary to keep the iscsi_put_conn() in the
> iscsi_iter_destroy_conn_fn(), otherwise, there will be memory leak
> problem.
>
I must have thought we did a unregister instead of remove for
some reason. Thanks for catching this.
Reviewed-by: Mike Christie <michael....@oracle.com>
> This reverts commit c577ab7ba5f3bf9062db8a58b6e89d4fe370447e.
>
> The invocation of iscsi_put_conn() in iscsi_iter_destory_conn_fn() is used
> to free the initial reference counter of iscsi_cls_conn.
> For non-qla4xxx cases, the ->destroy_conn() callback (e.g.,
> iscsi_conn_teardown) will call iscsi_remove_conn() and iscsi_put_conn() to
> remove the connection from the children list of session and free the
> connection at last.
> However for qla4xxx, it is not the case. The ->destroy_conn() callback
> of qla4xxx will keep the connection in the session conn_list and doesn't
> use iscsi_put_conn() to free the initial reference counter. Therefore,
> it seems necessary to keep the iscsi_put_conn() in the
> iscsi_iter_destroy_conn_fn(), otherwise, there will be memory leak
> problem.
>
I must have thought we did a unregister instead of remove for
some reason. Thanks for catching this.
Reviewed-by: Mike Christie <michael....@oracle.com>
Martin K. Petersen
Jul 24, 2025, 9:21:54 PMJul 24
to Li Lingfeng, ldu...@suse.com, cle...@redhat.com, nja...@marvell.com, mrang...@marvell.com, GR-QLogic-Sto...@marvell.com, martin....@oracle.com, je...@linux.ibm.com, James.B...@hansenpartnership.com, michael....@oracle.com, open-...@googlegroups.com, linux...@vger.kernel.org, linux-...@vger.kernel.org, yuk...@huaweicloud.com, hou...@huawei.com, yi.z...@huawei.com, yang...@huawei.com, chengz...@huawei.com, lilin...@huaweicloud.com
Li,
> This reverts commit c577ab7ba5f3bf9062db8a58b6e89d4fe370447e.
Applied to 6.17/scsi-staging, thanks!
--
Martin K. Petersen
Reply all
Reply to author
Forward
0 new messages