staging: lustre: ptlrpc: no need to reassign mbits for replay

It's not necessary reassgin & re-adjust rq_mbits for replay
request in ptlrpc_set_bulk_mbits(), they all must have already
been correctly assigned before.

Such unecessary reassign could make the first matchbit not
PTLRPC_BULK_OPS_MASK aligned, that'll trigger LASSERT in
ptlrpc_register_bulk():

- ptlrpc_set_bulk_mbits() is called when first time sending
  request, rq_mbits is set as xid, which is BULK_OPS aligned;

- ptlrpc_set_bulk_mbits() continue to adjust the mbits for
  multi-bulk RPC, rq_mbits is not aligned anymore, then rq_xid
  is changed accordingly if client is connecting to an old
  server, so rq_xid became unaligned too;

- The request is replayed, ptlrpc_set_bulk_mbits() reassign
  the rq_mbits as rq_xid, which isn't aligned already, but
  ptlrpc_register_bulk() still assumes this value as the
  first matchbits and LASSERT it's BULK_OPS aligned.

Signed-off-by: Niu Yawei <yawei.niu@intel.com>
Intel-bug-id: https://jira.hpdd.intel.com/browse/LU-6808
Reviewed-on: http://review.whamcloud.com/23048
Reviewed-by: Fan Yong <fan.yong@intel.com>
Reviewed-by: Alex Zhuravlev <alexey.zhuravlev@intel.com>
Reviewed-by: Oleg Drokin <oleg.drokin@intel.com>
Signed-off-by: James Simmons <jsimmons@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/drivers/staging/lustre/lustre/ptlrpc/client.c b/drivers/staging/lustre/lustre/ptlrpc/client.c
index 1c7779215eed4bb781070da200cf5058d9d47e18..1247686f1122dfc66d9ddf2e01d5f3df6a383909 100644 (file)
--- a/drivers/staging/lustre/lustre/ptlrpc/client.c
+++ b/drivers/staging/lustre/lustre/ptlrpc/client.c
@@ -3116,13 +3116,20 @@ void ptlrpc_set_bulk_mbits(struct ptlrpc_request *req)
 
        LASSERT(bd);
 
-       if (!req->rq_resend) {
-               /* this request has a new xid, just use it as bulk matchbits */
-               req->rq_mbits = req->rq_xid;
-
-       } else { /* needs to generate a new matchbits for resend */
+       /*
+        * Generate new matchbits for all resend requests, including
+        * resend replay.
+        */
+       if (req->rq_resend) {
                u64 old_mbits = req->rq_mbits;
 
+               /*
+                * First time resend on -EINPROGRESS will generate new xid,
+                * so we can actually use the rq_xid as rq_mbits in such case,
+                * however, it's bit hard to distinguish such resend with a
+                * 'resend for the -EINPROGRESS resend'. To make it simple,
+                * we opt to generate mbits for all resend cases.
+                */
                if ((bd->bd_import->imp_connect_data.ocd_connect_flags &
                     OBD_CONNECT_BULK_MBITS)) {
                        req->rq_mbits = ptlrpc_next_xid();
@@ -3131,12 +3138,21 @@ void ptlrpc_set_bulk_mbits(struct ptlrpc_request *req)
                        spin_lock(&req->rq_import->imp_lock);
                        list_del_init(&req->rq_unreplied_list);
                        ptlrpc_assign_next_xid_nolock(req);
-                       req->rq_mbits = req->rq_xid;
                        spin_unlock(&req->rq_import->imp_lock);
+                       req->rq_mbits = req->rq_xid;
                }
 
                CDEBUG(D_HA, "resend bulk old x%llu new x%llu\n",
                       old_mbits, req->rq_mbits);
+       } else if (!(lustre_msg_get_flags(req->rq_reqmsg) & MSG_REPLAY)) {
+               /* Request being sent first time, use xid as matchbits. */
+               req->rq_mbits = req->rq_xid;
+       } else {
+               /*
+                * Replay request, xid and matchbits have already been
+                * correctly assigned.
+                */
+               return;
        }
 
        /*