[PATCH ibacm 4/8] acm.c: Fix endian of transaction ID

[PATCH ibacm 4/8] acm.c: Fix endian of transaction ID

[Hal Rosenstock]

It should be in network rather than host endian.

Signed-off-by: Hal Rosenstock <hal-VPRAkNaXOzVWk0Htik3J/w@public.gmane.org>
---
 src/acm.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/acm.c b/src/acm.c
index 7de5873..d68746b 100644
--- a/src/acm.c
+++ b/src/acm.c
@@ -839,7 +839,7 @@ static void acm_init_path_query(struct ib_sa_mad *mad)
 	mad->mgmt_class = IB_MGMT_CLASS_SA;
 	mad->class_version = 2;
 	mad->method = IB_METHOD_GET;
-	mad->tid = (uint64_t) atomic_inc(&tid);
+	mad->tid = htonll((uint64_t) atomic_inc(&tid));
 	mad->attr_id = IB_SA_ATTR_PATH_REC;
 }
 
@@ -1483,7 +1483,7 @@ static void acm_init_join(struct ib_sa_mad *mad, union ibv_gid *port_gid,
 	mad->mgmt_class = IB_MGMT_CLASS_SA;
 	mad->class_version = 2;
 	mad->method = IB_METHOD_SET;
-	mad->tid = (uint64_t) atomic_inc(&tid);
+	mad->tid = htonll((uint64_t) atomic_inc(&tid));
 	mad->attr_id = IB_SA_ATTR_MC_MEMBER_REC;
 	mad->comp_mask =
 		IB_COMP_MASK_MC_MGID | IB_COMP_MASK_MC_PORT_GID |
@@ -1953,7 +1953,7 @@ acm_send_resolve(struct acm_ep *ep, struct acm_dest *dest,
 	mad->class_version = 1;
 	mad->method = IB_METHOD_GET;
 	mad->control = ACM_CTRL_RESOLVE;
-	mad->tid = (uint64_t) atomic_inc(&tid);
+	mad->tid = htonll((uint64_t) atomic_inc(&tid));
 
 	rec = (struct acm_resolve_rec *) mad->data;
 	rec->src_type = (uint8_t) saddr->type;
-- 
1.7.8.2

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

RE: [PATCH ibacm 4/8] acm.c: Fix endian of transaction ID

[Hefty, Sean]

> It should be in network rather than host endian.

The TID is basically an abstract 64-bit handle.  I don't see that it matters what value is used, provided that the sender use it consistently.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH ibacm 4/8] acm.c: Fix endian of transaction ID

[Hal Rosenstock]

On 8/7/2013 1:29 PM, Hefty, Sean wrote:
>> It should be in network rather than host endian.
> 
> The TID is basically an abstract 64-bit handle.  I don't see that it matters what value is used, 
> provided that the sender use it consistently.

With little endian machines, the current code puts the TID in the high
bits only to be overwritten by the agent ID so the TID is effectively 0.
That being said, it works either way since currently there is only 1
outstanding query. It's a matter of better TID usage/correctness and
debug when looking at OpenSM logs as to TIDs.

-- Hal

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html