[LU-13229] Flock unlock request resend caused a memory leak.

Thu, 5 Mar 2020 22:48:10 +0000

~~LU-12828~~ fix introduce a regression caused a request leak in the resend case.

@@ -922,6 +932,7 @@ resend:
                LASSERTF(einfo->ei_type == LDLM_FLOCK, “lock type %d\n”,
                         einfo->ei_type);
                res_id.name[3] = LDLM_FLOCK;
+               req = ldlm_enqueue_pack(exp, 0);
...
@@ -973,6 +971,7 @@ resend:
        rc = ldlm_cli_enqueue(exp, &req, einfo, &res_id, policy, &flags, NULL,
                              0, lvb_type, lockh, 0);
+
        if (!it) {
                /* For flock requests we immediatelly return without further
                   delay and let caller deal with the rest, since rest of
@@ -986,12 +985,10 @@ resend:
                    (einfo->ei_type == LDLM_FLOCK) &&
                    (einfo->ei_mode == LCK_NL))
                        goto resend;
+               ptlrpc_req_finished(req);
                RETURN(rc);
        }

So new request will allocated on resend case, but it don't freed in unlock resend case.

Whamcloud Community JIRA

[LU-13229] Flock unlock request resend caused a memory leak.