SVCRDMA: Use only 1 RDMA read scatter entry for iWARP adapters
[sfrench/cifs-2.6.git] / net / sunrpc / xprtrdma / svc_rdma_recvfrom.c
1 /*
2  * Copyright (c) 2005-2006 Network Appliance, Inc. All rights reserved.
3  *
4  * This software is available to you under a choice of one of two
5  * licenses.  You may choose to be licensed under the terms of the GNU
6  * General Public License (GPL) Version 2, available from the file
7  * COPYING in the main directory of this source tree, or the BSD-type
8  * license below:
9  *
10  * Redistribution and use in source and binary forms, with or without
11  * modification, are permitted provided that the following conditions
12  * are met:
13  *
14  *      Redistributions of source code must retain the above copyright
15  *      notice, this list of conditions and the following disclaimer.
16  *
17  *      Redistributions in binary form must reproduce the above
18  *      copyright notice, this list of conditions and the following
19  *      disclaimer in the documentation and/or other materials provided
20  *      with the distribution.
21  *
22  *      Neither the name of the Network Appliance, Inc. nor the names of
23  *      its contributors may be used to endorse or promote products
24  *      derived from this software without specific prior written
25  *      permission.
26  *
27  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
28  * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
29  * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
30  * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
31  * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
32  * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
33  * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
34  * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
35  * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
36  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
37  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
38  *
39  * Author: Tom Tucker <tom@opengridcomputing.com>
40  */
41
42 #include <linux/sunrpc/debug.h>
43 #include <linux/sunrpc/rpc_rdma.h>
44 #include <linux/spinlock.h>
45 #include <asm/unaligned.h>
46 #include <rdma/ib_verbs.h>
47 #include <rdma/rdma_cm.h>
48 #include <linux/sunrpc/svc_rdma.h>
49
50 #define RPCDBG_FACILITY RPCDBG_SVCXPRT
51
52 /*
53  * Replace the pages in the rq_argpages array with the pages from the SGE in
54  * the RDMA_RECV completion. The SGL should contain full pages up until the
55  * last one.
56  */
57 static void rdma_build_arg_xdr(struct svc_rqst *rqstp,
58                                struct svc_rdma_op_ctxt *ctxt,
59                                u32 byte_count)
60 {
61         struct page *page;
62         u32 bc;
63         int sge_no;
64
65         /* Swap the page in the SGE with the page in argpages */
66         page = ctxt->pages[0];
67         put_page(rqstp->rq_pages[0]);
68         rqstp->rq_pages[0] = page;
69
70         /* Set up the XDR head */
71         rqstp->rq_arg.head[0].iov_base = page_address(page);
72         rqstp->rq_arg.head[0].iov_len = min(byte_count, ctxt->sge[0].length);
73         rqstp->rq_arg.len = byte_count;
74         rqstp->rq_arg.buflen = byte_count;
75
76         /* Compute bytes past head in the SGL */
77         bc = byte_count - rqstp->rq_arg.head[0].iov_len;
78
79         /* If data remains, store it in the pagelist */
80         rqstp->rq_arg.page_len = bc;
81         rqstp->rq_arg.page_base = 0;
82         rqstp->rq_arg.pages = &rqstp->rq_pages[1];
83         sge_no = 1;
84         while (bc && sge_no < ctxt->count) {
85                 page = ctxt->pages[sge_no];
86                 put_page(rqstp->rq_pages[sge_no]);
87                 rqstp->rq_pages[sge_no] = page;
88                 bc -= min(bc, ctxt->sge[sge_no].length);
89                 rqstp->rq_arg.buflen += ctxt->sge[sge_no].length;
90                 sge_no++;
91         }
92         rqstp->rq_respages = &rqstp->rq_pages[sge_no];
93
94         /* We should never run out of SGE because the limit is defined to
95          * support the max allowed RPC data length
96          */
97         BUG_ON(bc && (sge_no == ctxt->count));
98         BUG_ON((rqstp->rq_arg.head[0].iov_len + rqstp->rq_arg.page_len)
99                != byte_count);
100         BUG_ON(rqstp->rq_arg.len != byte_count);
101
102         /* If not all pages were used from the SGL, free the remaining ones */
103         bc = sge_no;
104         while (sge_no < ctxt->count) {
105                 page = ctxt->pages[sge_no++];
106                 put_page(page);
107         }
108         ctxt->count = bc;
109
110         /* Set up tail */
111         rqstp->rq_arg.tail[0].iov_base = NULL;
112         rqstp->rq_arg.tail[0].iov_len = 0;
113 }
114
115 struct chunk_sge {
116         int start;              /* sge no for this chunk */
117         int count;              /* sge count for this chunk */
118 };
119
120 /* Encode a read-chunk-list as an array of IB SGE
121  *
122  * Assumptions:
123  * - chunk[0]->position points to pages[0] at an offset of 0
124  * - pages[] is not physically or virtually contigous and consists of
125  *   PAGE_SIZE elements.
126  *
127  * Output:
128  * - sge array pointing into pages[] array.
129  * - chunk_sge array specifying sge index and count for each
130  *   chunk in the read list
131  *
132  */
133 static int rdma_rcl_to_sge(struct svcxprt_rdma *xprt,
134                            struct svc_rqst *rqstp,
135                            struct svc_rdma_op_ctxt *head,
136                            struct rpcrdma_msg *rmsgp,
137                            struct ib_sge *sge,
138                            struct chunk_sge *ch_sge_ary,
139                            int ch_count,
140                            int byte_count)
141 {
142         int sge_no;
143         int sge_bytes;
144         int page_off;
145         int page_no;
146         int ch_bytes;
147         int ch_no;
148         struct rpcrdma_read_chunk *ch;
149
150         sge_no = 0;
151         page_no = 0;
152         page_off = 0;
153         ch = (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0];
154         ch_no = 0;
155         ch_bytes = ch->rc_target.rs_length;
156         head->arg.head[0] = rqstp->rq_arg.head[0];
157         head->arg.tail[0] = rqstp->rq_arg.tail[0];
158         head->arg.pages = &head->pages[head->count];
159         head->sge[0].length = head->count; /* save count of hdr pages */
160         head->arg.page_base = 0;
161         head->arg.page_len = ch_bytes;
162         head->arg.len = rqstp->rq_arg.len + ch_bytes;
163         head->arg.buflen = rqstp->rq_arg.buflen + ch_bytes;
164         head->count++;
165         ch_sge_ary[0].start = 0;
166         while (byte_count) {
167                 sge_bytes = min_t(int, PAGE_SIZE-page_off, ch_bytes);
168                 sge[sge_no].addr =
169                         ib_dma_map_page(xprt->sc_cm_id->device,
170                                         rqstp->rq_arg.pages[page_no],
171                                         page_off, sge_bytes,
172                                         DMA_FROM_DEVICE);
173                 sge[sge_no].length = sge_bytes;
174                 sge[sge_no].lkey = xprt->sc_phys_mr->lkey;
175                 /*
176                  * Don't bump head->count here because the same page
177                  * may be used by multiple SGE.
178                  */
179                 head->arg.pages[page_no] = rqstp->rq_arg.pages[page_no];
180                 rqstp->rq_respages = &rqstp->rq_arg.pages[page_no+1];
181
182                 byte_count -= sge_bytes;
183                 ch_bytes -= sge_bytes;
184                 sge_no++;
185                 /*
186                  * If all bytes for this chunk have been mapped to an
187                  * SGE, move to the next SGE
188                  */
189                 if (ch_bytes == 0) {
190                         ch_sge_ary[ch_no].count =
191                                 sge_no - ch_sge_ary[ch_no].start;
192                         ch_no++;
193                         ch++;
194                         ch_sge_ary[ch_no].start = sge_no;
195                         ch_bytes = ch->rc_target.rs_length;
196                         /* If bytes remaining account for next chunk */
197                         if (byte_count) {
198                                 head->arg.page_len += ch_bytes;
199                                 head->arg.len += ch_bytes;
200                                 head->arg.buflen += ch_bytes;
201                         }
202                 }
203                 /*
204                  * If this SGE consumed all of the page, move to the
205                  * next page
206                  */
207                 if ((sge_bytes + page_off) == PAGE_SIZE) {
208                         page_no++;
209                         page_off = 0;
210                         /*
211                          * If there are still bytes left to map, bump
212                          * the page count
213                          */
214                         if (byte_count)
215                                 head->count++;
216                 } else
217                         page_off += sge_bytes;
218         }
219         BUG_ON(byte_count != 0);
220         return sge_no;
221 }
222
223 static void rdma_set_ctxt_sge(struct svc_rdma_op_ctxt *ctxt,
224                               struct ib_sge *sge,
225                               u64 *sgl_offset,
226                               int count)
227 {
228         int i;
229
230         ctxt->count = count;
231         for (i = 0; i < count; i++) {
232                 ctxt->sge[i].addr = sge[i].addr;
233                 ctxt->sge[i].length = sge[i].length;
234                 *sgl_offset = *sgl_offset + sge[i].length;
235         }
236 }
237
238 static int rdma_read_max_sge(struct svcxprt_rdma *xprt, int sge_count)
239 {
240         if ((RDMA_TRANSPORT_IWARP ==
241              rdma_node_get_transport(xprt->sc_cm_id->
242                                      device->node_type))
243             && sge_count > 1)
244                 return 1;
245         else
246                 return min_t(int, sge_count, xprt->sc_max_sge);
247 }
248
249 /*
250  * Use RDMA_READ to read data from the advertised client buffer into the
251  * XDR stream starting at rq_arg.head[0].iov_base.
252  * Each chunk in the array
253  * contains the following fields:
254  * discrim      - '1', This isn't used for data placement
255  * position     - The xdr stream offset (the same for every chunk)
256  * handle       - RMR for client memory region
257  * length       - data transfer length
258  * offset       - 64 bit tagged offset in remote memory region
259  *
260  * On our side, we need to read into a pagelist. The first page immediately
261  * follows the RPC header.
262  *
263  * This function returns 1 to indicate success. The data is not yet in
264  * the pagelist and therefore the RPC request must be deferred. The
265  * I/O completion will enqueue the transport again and
266  * svc_rdma_recvfrom will complete the request.
267  *
268  * NOTE: The ctxt must not be touched after the last WR has been posted
269  * because the I/O completion processing may occur on another
270  * processor and free / modify the context. Ne touche pas!
271  */
272 static int rdma_read_xdr(struct svcxprt_rdma *xprt,
273                          struct rpcrdma_msg *rmsgp,
274                          struct svc_rqst *rqstp,
275                          struct svc_rdma_op_ctxt *hdr_ctxt)
276 {
277         struct ib_send_wr read_wr;
278         int err = 0;
279         int ch_no;
280         struct ib_sge *sge;
281         int ch_count;
282         int byte_count;
283         int sge_count;
284         u64 sgl_offset;
285         struct rpcrdma_read_chunk *ch;
286         struct svc_rdma_op_ctxt *ctxt = NULL;
287         struct svc_rdma_op_ctxt *head;
288         struct svc_rdma_op_ctxt *tmp_sge_ctxt;
289         struct svc_rdma_op_ctxt *tmp_ch_ctxt;
290         struct chunk_sge *ch_sge_ary;
291
292         /* If no read list is present, return 0 */
293         ch = svc_rdma_get_read_chunk(rmsgp);
294         if (!ch)
295                 return 0;
296
297         /* Allocate temporary contexts to keep SGE */
298         BUG_ON(sizeof(struct ib_sge) < sizeof(struct chunk_sge));
299         tmp_sge_ctxt = svc_rdma_get_context(xprt);
300         sge = tmp_sge_ctxt->sge;
301         tmp_ch_ctxt = svc_rdma_get_context(xprt);
302         ch_sge_ary = (struct chunk_sge *)tmp_ch_ctxt->sge;
303
304         svc_rdma_rcl_chunk_counts(ch, &ch_count, &byte_count);
305         sge_count = rdma_rcl_to_sge(xprt, rqstp, hdr_ctxt, rmsgp,
306                                     sge, ch_sge_ary,
307                                     ch_count, byte_count);
308         head = svc_rdma_get_context(xprt);
309         sgl_offset = 0;
310         ch_no = 0;
311
312         for (ch = (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0];
313              ch->rc_discrim != 0; ch++, ch_no++) {
314 next_sge:
315                 if (!ctxt)
316                         ctxt = head;
317                 else {
318                         ctxt->next = svc_rdma_get_context(xprt);
319                         ctxt = ctxt->next;
320                 }
321                 ctxt->next = NULL;
322                 ctxt->direction = DMA_FROM_DEVICE;
323                 clear_bit(RDMACTXT_F_READ_DONE, &ctxt->flags);
324                 clear_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags);
325                 if ((ch+1)->rc_discrim == 0) {
326                         /*
327                          * Checked in sq_cq_reap to see if we need to
328                          * be enqueued
329                          */
330                         set_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags);
331                         ctxt->next = hdr_ctxt;
332                         hdr_ctxt->next = head;
333                 }
334
335                 /* Prepare READ WR */
336                 memset(&read_wr, 0, sizeof read_wr);
337                 ctxt->wr_op = IB_WR_RDMA_READ;
338                 read_wr.wr_id = (unsigned long)ctxt;
339                 read_wr.opcode = IB_WR_RDMA_READ;
340                 read_wr.send_flags = IB_SEND_SIGNALED;
341                 read_wr.wr.rdma.rkey = ch->rc_target.rs_handle;
342                 read_wr.wr.rdma.remote_addr =
343                         get_unaligned(&(ch->rc_target.rs_offset)) +
344                         sgl_offset;
345                 read_wr.sg_list = &sge[ch_sge_ary[ch_no].start];
346                 read_wr.num_sge =
347                         rdma_read_max_sge(xprt, ch_sge_ary[ch_no].count);
348                 rdma_set_ctxt_sge(ctxt, &sge[ch_sge_ary[ch_no].start],
349                                   &sgl_offset,
350                                   read_wr.num_sge);
351
352                 /* Post the read */
353                 err = svc_rdma_send(xprt, &read_wr);
354                 if (err) {
355                         printk(KERN_ERR "svcrdma: Error posting send = %d\n",
356                                err);
357                         /*
358                          * Break the circular list so free knows when
359                          * to stop if the error happened to occur on
360                          * the last read
361                          */
362                         ctxt->next = NULL;
363                         goto out;
364                 }
365                 atomic_inc(&rdma_stat_read);
366
367                 if (read_wr.num_sge < ch_sge_ary[ch_no].count) {
368                         ch_sge_ary[ch_no].count -= read_wr.num_sge;
369                         ch_sge_ary[ch_no].start += read_wr.num_sge;
370                         goto next_sge;
371                 }
372                 sgl_offset = 0;
373                 err = 0;
374         }
375
376  out:
377         svc_rdma_put_context(tmp_sge_ctxt, 0);
378         svc_rdma_put_context(tmp_ch_ctxt, 0);
379
380         /* Detach arg pages. svc_recv will replenish them */
381         for (ch_no = 0; &rqstp->rq_pages[ch_no] < rqstp->rq_respages; ch_no++)
382                 rqstp->rq_pages[ch_no] = NULL;
383
384         /*
385          * Detach res pages. svc_release must see a resused count of
386          * zero or it will attempt to put them.
387          */
388         while (rqstp->rq_resused)
389                 rqstp->rq_respages[--rqstp->rq_resused] = NULL;
390
391         if (err) {
392                 printk(KERN_ERR "svcrdma : RDMA_READ error = %d\n", err);
393                 set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags);
394                 /* Free the linked list of read contexts */
395                 while (head != NULL) {
396                         ctxt = head->next;
397                         svc_rdma_put_context(head, 1);
398                         head = ctxt;
399                 }
400                 return 0;
401         }
402
403         return 1;
404 }
405
406 static int rdma_read_complete(struct svc_rqst *rqstp,
407                               struct svc_rdma_op_ctxt *data)
408 {
409         struct svc_rdma_op_ctxt *head = data->next;
410         int page_no;
411         int ret;
412
413         BUG_ON(!head);
414
415         /* Copy RPC pages */
416         for (page_no = 0; page_no < head->count; page_no++) {
417                 put_page(rqstp->rq_pages[page_no]);
418                 rqstp->rq_pages[page_no] = head->pages[page_no];
419         }
420         /* Point rq_arg.pages past header */
421         rqstp->rq_arg.pages = &rqstp->rq_pages[head->sge[0].length];
422         rqstp->rq_arg.page_len = head->arg.page_len;
423         rqstp->rq_arg.page_base = head->arg.page_base;
424
425         /* rq_respages starts after the last arg page */
426         rqstp->rq_respages = &rqstp->rq_arg.pages[page_no];
427         rqstp->rq_resused = 0;
428
429         /* Rebuild rq_arg head and tail. */
430         rqstp->rq_arg.head[0] = head->arg.head[0];
431         rqstp->rq_arg.tail[0] = head->arg.tail[0];
432         rqstp->rq_arg.len = head->arg.len;
433         rqstp->rq_arg.buflen = head->arg.buflen;
434
435         /* XXX: What should this be? */
436         rqstp->rq_prot = IPPROTO_MAX;
437
438         /*
439          * Free the contexts we used to build the RDMA_READ. We have
440          * to be careful here because the context list uses the same
441          * next pointer used to chain the contexts associated with the
442          * RDMA_READ
443          */
444         data->next = NULL;      /* terminate circular list */
445         do {
446                 data = head->next;
447                 svc_rdma_put_context(head, 0);
448                 head = data;
449         } while (head != NULL);
450
451         ret = rqstp->rq_arg.head[0].iov_len
452                 + rqstp->rq_arg.page_len
453                 + rqstp->rq_arg.tail[0].iov_len;
454         dprintk("svcrdma: deferred read ret=%d, rq_arg.len =%d, "
455                 "rq_arg.head[0].iov_base=%p, rq_arg.head[0].iov_len = %zd\n",
456                 ret, rqstp->rq_arg.len, rqstp->rq_arg.head[0].iov_base,
457                 rqstp->rq_arg.head[0].iov_len);
458
459         /* Indicate that we've consumed an RQ credit */
460         rqstp->rq_xprt_ctxt = rqstp->rq_xprt;
461         svc_xprt_received(rqstp->rq_xprt);
462         return ret;
463 }
464
465 /*
466  * Set up the rqstp thread context to point to the RQ buffer. If
467  * necessary, pull additional data from the client with an RDMA_READ
468  * request.
469  */
470 int svc_rdma_recvfrom(struct svc_rqst *rqstp)
471 {
472         struct svc_xprt *xprt = rqstp->rq_xprt;
473         struct svcxprt_rdma *rdma_xprt =
474                 container_of(xprt, struct svcxprt_rdma, sc_xprt);
475         struct svc_rdma_op_ctxt *ctxt = NULL;
476         struct rpcrdma_msg *rmsgp;
477         int ret = 0;
478         int len;
479
480         dprintk("svcrdma: rqstp=%p\n", rqstp);
481
482         /*
483          * The rq_xprt_ctxt indicates if we've consumed an RQ credit
484          * or not. It is used in the rdma xpo_release_rqst function to
485          * determine whether or not to return an RQ WQE to the RQ.
486          */
487         rqstp->rq_xprt_ctxt = NULL;
488
489         spin_lock_bh(&rdma_xprt->sc_read_complete_lock);
490         if (!list_empty(&rdma_xprt->sc_read_complete_q)) {
491                 ctxt = list_entry(rdma_xprt->sc_read_complete_q.next,
492                                   struct svc_rdma_op_ctxt,
493                                   dto_q);
494                 list_del_init(&ctxt->dto_q);
495         }
496         spin_unlock_bh(&rdma_xprt->sc_read_complete_lock);
497         if (ctxt)
498                 return rdma_read_complete(rqstp, ctxt);
499
500         spin_lock_bh(&rdma_xprt->sc_rq_dto_lock);
501         if (!list_empty(&rdma_xprt->sc_rq_dto_q)) {
502                 ctxt = list_entry(rdma_xprt->sc_rq_dto_q.next,
503                                   struct svc_rdma_op_ctxt,
504                                   dto_q);
505                 list_del_init(&ctxt->dto_q);
506         } else {
507                 atomic_inc(&rdma_stat_rq_starve);
508                 clear_bit(XPT_DATA, &xprt->xpt_flags);
509                 ctxt = NULL;
510         }
511         spin_unlock_bh(&rdma_xprt->sc_rq_dto_lock);
512         if (!ctxt) {
513                 /* This is the EAGAIN path. The svc_recv routine will
514                  * return -EAGAIN, the nfsd thread will go to call into
515                  * svc_recv again and we shouldn't be on the active
516                  * transport list
517                  */
518                 if (test_bit(XPT_CLOSE, &xprt->xpt_flags))
519                         goto close_out;
520
521                 BUG_ON(ret);
522                 goto out;
523         }
524         dprintk("svcrdma: processing ctxt=%p on xprt=%p, rqstp=%p, status=%d\n",
525                 ctxt, rdma_xprt, rqstp, ctxt->wc_status);
526         BUG_ON(ctxt->wc_status != IB_WC_SUCCESS);
527         atomic_inc(&rdma_stat_recv);
528
529         /* Build up the XDR from the receive buffers. */
530         rdma_build_arg_xdr(rqstp, ctxt, ctxt->byte_len);
531
532         /* Decode the RDMA header. */
533         len = svc_rdma_xdr_decode_req(&rmsgp, rqstp);
534         rqstp->rq_xprt_hlen = len;
535
536         /* If the request is invalid, reply with an error */
537         if (len < 0) {
538                 if (len == -ENOSYS)
539                         (void)svc_rdma_send_error(rdma_xprt, rmsgp, ERR_VERS);
540                 goto close_out;
541         }
542
543         /* Read read-list data. If we would need to wait, defer
544          * it. Not that in this case, we don't return the RQ credit
545          * until after the read completes.
546          */
547         if (rdma_read_xdr(rdma_xprt, rmsgp, rqstp, ctxt)) {
548                 svc_xprt_received(xprt);
549                 return 0;
550         }
551
552         /* Indicate we've consumed an RQ credit */
553         rqstp->rq_xprt_ctxt = rqstp->rq_xprt;
554
555         ret = rqstp->rq_arg.head[0].iov_len
556                 + rqstp->rq_arg.page_len
557                 + rqstp->rq_arg.tail[0].iov_len;
558         svc_rdma_put_context(ctxt, 0);
559  out:
560         dprintk("svcrdma: ret = %d, rq_arg.len =%d, "
561                 "rq_arg.head[0].iov_base=%p, rq_arg.head[0].iov_len = %zd\n",
562                 ret, rqstp->rq_arg.len,
563                 rqstp->rq_arg.head[0].iov_base,
564                 rqstp->rq_arg.head[0].iov_len);
565         rqstp->rq_prot = IPPROTO_MAX;
566         svc_xprt_copy_addrs(rqstp, xprt);
567         svc_xprt_received(xprt);
568         return ret;
569
570  close_out:
571         if (ctxt) {
572                 svc_rdma_put_context(ctxt, 1);
573                 /* Indicate we've consumed an RQ credit */
574                 rqstp->rq_xprt_ctxt = rqstp->rq_xprt;
575         }
576         dprintk("svcrdma: transport %p is closing\n", xprt);
577         /*
578          * Set the close bit and enqueue it. svc_recv will see the
579          * close bit and call svc_xprt_delete
580          */
581         set_bit(XPT_CLOSE, &xprt->xpt_flags);
582         svc_xprt_received(xprt);
583         return 0;
584 }