Commit 514349dc authored by Trond Myklebust's avatar Trond Myklebust

[PATCH] RPC over UDP congestion control updates [5/8]

Clean up the Van Jacobson network congestion control code.
parent c6b43f23
......@@ -19,7 +19,7 @@
* The transport code maintains an estimate on the maximum number of out-
* standing RPC requests, using a smoothed version of the congestion
* avoidance implemented in 44BSD. This is basically the Van Jacobson
* slow start algorithm: If a retransmit occurs, the congestion window is
* congestion algorithm: If a retransmit occurs, the congestion window is
* halved; otherwise, it is incremented by 1/cwnd when
*
* - a reply is received and
......@@ -32,15 +32,13 @@
* Note: on machines with low memory we should probably use a smaller
* MAXREQS value: At 32 outstanding reqs with 8 megs of RAM, fragment
* reassembly will frequently run out of memory.
* Come Linux 2.3, we'll handle fragments directly.
*/
#define RPC_MAXCONG 16
#define RPC_MAXREQS (RPC_MAXCONG + 1)
#define RPC_CWNDSCALE 256
#define RPC_MAXCWND (RPC_MAXCONG * RPC_CWNDSCALE)
#define RPC_INITCWND RPC_CWNDSCALE
#define RPCXPRT_CONGESTED(xprt) \
((xprt)->cong >= (xprt)->cwnd)
#define RPCXPRT_CONGESTED(xprt) ((xprt)->cong >= (xprt)->cwnd)
/* Default timeout values */
#define RPC_MAX_UDP_TIMEOUT (60*HZ)
......@@ -83,6 +81,7 @@ struct rpc_rqst {
struct rpc_task * rq_task; /* RPC task data */
__u32 rq_xid; /* request XID */
struct rpc_rqst * rq_next; /* free list */
int rq_cong; /* has incremented xprt->cong */
int rq_received; /* receive completed */
struct list_head rq_list;
......
......@@ -89,6 +89,7 @@ static void xprt_disconnect(struct rpc_xprt *);
static void xprt_reconn_status(struct rpc_task *task);
static struct socket *xprt_create_socket(int, struct rpc_timeout *);
static int xprt_bind_socket(struct rpc_xprt *, struct socket *);
static int __xprt_get_cong(struct rpc_xprt *, struct rpc_task *);
#ifdef RPC_DEBUG_DATA
/*
......@@ -253,6 +254,40 @@ xprt_sendmsg(struct rpc_xprt *xprt, struct rpc_rqst *req)
return result;
}
/*
* Van Jacobson congestion avoidance. Check if the congestion window
* overflowed. Put the task to sleep if this is the case.
*/
static int
__xprt_get_cong(struct rpc_xprt *xprt, struct rpc_task *task)
{
struct rpc_rqst *req = task->tk_rqstp;
if (req->rq_cong)
return 1;
dprintk("RPC: %4d xprt_cwnd_limited cong = %ld cwnd = %ld\n",
task->tk_pid, xprt->cong, xprt->cwnd);
if (RPCXPRT_CONGESTED(xprt))
return 0;
req->rq_cong = 1;
xprt->cong += RPC_CWNDSCALE;
return 1;
}
/*
* Adjust the congestion window, and wake up the next task
* that has been sleeping due to congestion
*/
static void
__xprt_put_cong(struct rpc_xprt *xprt, struct rpc_rqst *req)
{
if (!req->rq_cong)
return;
req->rq_cong = 0;
xprt->cong -= RPC_CWNDSCALE;
__xprt_lock_write_next(xprt);
}
/*
* Adjust RPC congestion window
* We use a time-smoothed congestion estimator to avoid heavy oscillation.
......@@ -1146,8 +1181,6 @@ xprt_reserve(struct rpc_task *task)
if (task->tk_rqstp)
return 0;
dprintk("RPC: %4d xprt_reserve cong = %ld cwnd = %ld\n",
task->tk_pid, xprt->cong, xprt->cwnd);
spin_lock_bh(&xprt->xprt_lock);
xprt_reserve_status(task);
if (task->tk_rqstp) {
......@@ -1181,13 +1214,14 @@ xprt_reserve_status(struct rpc_task *task)
} else if (task->tk_rqstp) {
/* We've already been given a request slot: NOP */
} else {
if (RPCXPRT_CONGESTED(xprt) || !(req = xprt->free))
if (!(req = xprt->free))
goto out_nofree;
if (!(xprt->nocong || __xprt_get_cong(xprt, req)))
goto out_nofree;
/* OK: There's room for us. Grab a free slot and bump
* congestion value */
xprt->free = req->rq_next;
req->rq_next = NULL;
xprt->cong += RPC_CWNDSCALE;
task->tk_rqstp = req;
xprt_request_init(task, xprt);
......@@ -1252,9 +1286,7 @@ xprt_release(struct rpc_task *task)
req->rq_next = xprt->free;
xprt->free = req;
/* Decrease congestion value. */
xprt->cong -= RPC_CWNDSCALE;
__xprt_put_cong(xprt, req);
xprt_clear_backlog(xprt);
spin_unlock_bh(&xprt->xprt_lock);
}
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment