iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

xprtrdma: Prevent leak of rpcrdma_rep objects

Browse Source 
[ Upstream commit 07e10308ee5da8e6132e0b737ece1c99dd651fb6 ]

If a reply has been processed but the RPC is later retransmitted
anyway, the req->rl_reply field still contains the only pointer to
the old rpcrdma rep. When the next reply comes in, the reply handler
will stomp on the rl_reply field, leaking the old rep.

A trace event is added to capture such leaks.

This problem seems to be worsened by the restructuring of the RPC
Call path in v4.20. Fully addressing this issue will require at
least a re-architecture of the disconnect logic, which is not
appropriate during -rc.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
This commit is contained in:
Chuck Lever 2018-12-07 11:11:44 -05:00 committed by Greg Kroah-Hartman
parent d398807976
commit d95b814359
2 changed files with 32 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
include/trace/events/rpcrdma.h
View File

@ -917,6 +917,34 @@ TRACE_EVENT(xprtrdma_cb_setup,
DEFINE_CB_EVENT(xprtrdma_cb_call);
DEFINE_CB_EVENT(xprtrdma_cb_reply);
TRACE_EVENT(xprtrdma_leaked_rep,
TP_PROTO(
const struct rpc_rqst *rqst,
const struct rpcrdma_rep *rep
),
TP_ARGS(rqst, rep),
TP_STRUCT__entry(
__field(unsigned int, task_id)
__field(unsigned int, client_id)
__field(u32, xid)
__field(const void *, rep)
),
TP_fast_assign(
__entry->task_id = rqst->rq_task->tk_pid;
__entry->client_id = rqst->rq_task->tk_client->cl_clid;
__entry->xid = be32_to_cpu(rqst->rq_xid);
__entry->rep = rep;
),
TP_printk("task:%u@%u xid=0x%08x rep=%p",
__entry->task_id, __entry->client_id, __entry->xid,
__entry->rep
)
);
/**
** Server-side RPC/RDMA events
**/

4
net/sunrpc/xprtrdma/rpc_rdma.c
View File

@ -1360,6 +1360,10 @@ void rpcrdma_reply_handler(struct rpcrdma_rep *rep)
spin_unlock(&xprt->recv_lock);
req = rpcr_to_rdmar(rqst);
if (req->rl_reply) {
trace_xprtrdma_leaked_rep(rqst, req->rl_reply);
rpcrdma_recv_buffer_put(req->rl_reply);
}
req->rl_reply = rep;
rep->rr_rqst = rqst;
clear_bit(RPCRDMA_REQ_F_PENDING, &req->rl_flags);