On 11/15/22 5:58 PM, Jinpu Wang wrote:
On Tue, Nov 15, 2022 at 10:53 AM Jinpu Wang <jinpu.wang@xxxxxxxxx> wrote:
On Tue, Nov 15, 2022 at 10:39 AM Haris Iqbal <haris.iqbal@xxxxxxxxx> wrote:
On Sun, Nov 13, 2022 at 2:08 AM Guoqing Jiang <guoqing.jiang@xxxxxxxxx> wrote:
Seems the only relevant comment about recon_cnt is,
/*
* On every new session connections increase reconnect counter
* to avoid clashes with previous sessions not yet closed
* sessions on a server side.
*/
However, it is not clear how the recon_cnt avoid clashed at these places
in the commit since no where checks it.
It does seem redundant. This predates my time, so I don't know if
there was a change which removed the usage of this. I tried to search
in commit history, but couldn't.
@Jinpu Your thoughts?
Sorry for sending an empty reply.
I checked a bit, and also can find a reason why the recon_cnt was
added, it is not checked on the server side.
I think it's ok to remove it. But we can simply remove the field in
rtrs_msg_conn_req, commented below.
Signed-off-by: Guoqing Jiang <guoqing.jiang@xxxxxxxxx>
---
drivers/infiniband/ulp/rtrs/rtrs-clt.c | 8 --------
drivers/infiniband/ulp/rtrs/rtrs-pri.h | 3 ---
drivers/infiniband/ulp/rtrs/rtrs-srv.c | 7 +------
3 files changed, 1 insertion(+), 17 deletions(-)
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-clt.c b/drivers/infiniband/ulp/rtrs/rtrs-clt.c
index 5ffc170dae8c..dcc8c041a141 100644
--- a/drivers/infiniband/ulp/rtrs/rtrs-clt.c
+++ b/drivers/infiniband/ulp/rtrs/rtrs-clt.c
@@ -1802,7 +1802,6 @@ static int rtrs_rdma_route_resolved(struct rtrs_clt_con *con)
.version = cpu_to_le16(RTRS_PROTO_VER),
.cid = cpu_to_le16(con->c.cid),
.cid_num = cpu_to_le16(clt_path->s.con_num),
- .recon_cnt = cpu_to_le16(clt_path->s.recon_cnt),
};
msg.first_conn = clt_path->for_new_clt ? FIRST_CONN : 0;
uuid_copy(&msg.sess_uuid, &clt_path->s.uuid);
@@ -2336,13 +2335,6 @@ static int init_conns(struct rtrs_clt_path *clt_path)
unsigned int cid;
int err;
- /*
- * On every new session connections increase reconnect counter
- * to avoid clashes with previous sessions not yet closed
- * sessions on a server side.
- */
- clt_path->s.recon_cnt++;
-
/* Establish all RDMA connections */
for (cid = 0; cid < clt_path->s.con_num; cid++) {
err = create_con(clt_path, cid);
diff --git a/drivers/infiniband/ulp/rtrs/rtrs-pri.h b/drivers/infiniband/ulp/rtrs/rtrs-pri.h
index a2420eecaf5a..c4ddaeba1c59 100644
--- a/drivers/infiniband/ulp/rtrs/rtrs-pri.h
+++ b/drivers/infiniband/ulp/rtrs/rtrs-pri.h
@@ -109,7 +109,6 @@ struct rtrs_path {
struct rtrs_con **con;
unsigned int con_num;
unsigned int irq_con_num;
- unsigned int recon_cnt;
unsigned int signal_interval;
struct rtrs_ib_dev *dev;
int dev_ref;
@@ -177,7 +176,6 @@ struct rtrs_sg_desc {
* @version: RTRS protocol version
* @cid: Current connection id
* @cid_num: Number of connections per session
- * @recon_cnt: Reconnections counter
* @sess_uuid: UUID of a session (path)
* @paths_uuid: UUID of a group of sessions (paths)
*
@@ -196,7 +194,6 @@ struct rtrs_msg_conn_req {
__le16 version;
__le16 cid;
__le16 cid_num;
- __le16 recon_cnt;
uuid_t sess_uuid;
uuid_t paths_uuid;
u8 first_conn : 1;
We can remove it from protocol, this break compatibility with older kernel.
Can we bump up RTRS_PROTO_VER_MINOR for the compatibility? Otherwise
those structs are immutable forever.
Thanks,
Guoqing
