Commit | Line | Data |
---|---|---|
c06b540a | 1 | /* |
0bf48289 | 2 | * Copyright (c) 2014 Open Grid Computing, Inc. All rights reserved. |
c06b540a TT |
3 | * Copyright (c) 2005-2006 Network Appliance, Inc. All rights reserved. |
4 | * | |
5 | * This software is available to you under a choice of one of two | |
6 | * licenses. You may choose to be licensed under the terms of the GNU | |
7 | * General Public License (GPL) Version 2, available from the file | |
8 | * COPYING in the main directory of this source tree, or the BSD-type | |
9 | * license below: | |
10 | * | |
11 | * Redistribution and use in source and binary forms, with or without | |
12 | * modification, are permitted provided that the following conditions | |
13 | * are met: | |
14 | * | |
15 | * Redistributions of source code must retain the above copyright | |
16 | * notice, this list of conditions and the following disclaimer. | |
17 | * | |
18 | * Redistributions in binary form must reproduce the above | |
19 | * copyright notice, this list of conditions and the following | |
20 | * disclaimer in the documentation and/or other materials provided | |
21 | * with the distribution. | |
22 | * | |
23 | * Neither the name of the Network Appliance, Inc. nor the names of | |
24 | * its contributors may be used to endorse or promote products | |
25 | * derived from this software without specific prior written | |
26 | * permission. | |
27 | * | |
28 | * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS | |
29 | * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT | |
30 | * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR | |
31 | * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT | |
32 | * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, | |
33 | * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT | |
34 | * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, | |
35 | * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY | |
36 | * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT | |
37 | * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE | |
38 | * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. | |
39 | * | |
40 | * Author: Tom Tucker <tom@opengridcomputing.com> | |
41 | */ | |
42 | ||
43 | #include <linux/sunrpc/debug.h> | |
44 | #include <linux/sunrpc/rpc_rdma.h> | |
45 | #include <linux/spinlock.h> | |
46 | #include <asm/unaligned.h> | |
47 | #include <rdma/ib_verbs.h> | |
48 | #include <rdma/rdma_cm.h> | |
49 | #include <linux/sunrpc/svc_rdma.h> | |
50 | ||
51 | #define RPCDBG_FACILITY RPCDBG_SVCXPRT | |
52 | ||
afd566ea TT |
53 | static int map_xdr(struct svcxprt_rdma *xprt, |
54 | struct xdr_buf *xdr, | |
55 | struct svc_rdma_req_map *vec) | |
c06b540a | 56 | { |
c06b540a | 57 | int sge_no; |
c06b540a TT |
58 | u32 sge_bytes; |
59 | u32 page_bytes; | |
34d16e42 | 60 | u32 page_off; |
c06b540a TT |
61 | int page_no; |
62 | ||
3fe04ee9 CL |
63 | if (xdr->len != |
64 | (xdr->head[0].iov_len + xdr->page_len + xdr->tail[0].iov_len)) { | |
65 | pr_err("svcrdma: map_xdr: XDR buffer length error\n"); | |
66 | return -EIO; | |
67 | } | |
34d16e42 | 68 | |
c06b540a TT |
69 | /* Skip the first sge, this is for the RPCRDMA header */ |
70 | sge_no = 1; | |
71 | ||
72 | /* Head SGE */ | |
34d16e42 TT |
73 | vec->sge[sge_no].iov_base = xdr->head[0].iov_base; |
74 | vec->sge[sge_no].iov_len = xdr->head[0].iov_len; | |
c06b540a TT |
75 | sge_no++; |
76 | ||
77 | /* pages SGE */ | |
78 | page_no = 0; | |
79 | page_bytes = xdr->page_len; | |
80 | page_off = xdr->page_base; | |
34d16e42 TT |
81 | while (page_bytes) { |
82 | vec->sge[sge_no].iov_base = | |
83 | page_address(xdr->pages[page_no]) + page_off; | |
84 | sge_bytes = min_t(u32, page_bytes, (PAGE_SIZE - page_off)); | |
c06b540a | 85 | page_bytes -= sge_bytes; |
34d16e42 | 86 | vec->sge[sge_no].iov_len = sge_bytes; |
c06b540a TT |
87 | |
88 | sge_no++; | |
89 | page_no++; | |
90 | page_off = 0; /* reset for next time through loop */ | |
91 | } | |
92 | ||
93 | /* Tail SGE */ | |
34d16e42 TT |
94 | if (xdr->tail[0].iov_len) { |
95 | vec->sge[sge_no].iov_base = xdr->tail[0].iov_base; | |
96 | vec->sge[sge_no].iov_len = xdr->tail[0].iov_len; | |
c06b540a TT |
97 | sge_no++; |
98 | } | |
99 | ||
b1e1e158 | 100 | dprintk("svcrdma: map_xdr: sge_no %d page_no %d " |
2e3c230b | 101 | "page_base %u page_len %u head_len %zu tail_len %zu\n", |
b1e1e158 TT |
102 | sge_no, page_no, xdr->page_base, xdr->page_len, |
103 | xdr->head[0].iov_len, xdr->tail[0].iov_len); | |
104 | ||
34d16e42 | 105 | vec->count = sge_no; |
afd566ea | 106 | return 0; |
c06b540a TT |
107 | } |
108 | ||
b432e6b3 TT |
109 | static dma_addr_t dma_map_xdr(struct svcxprt_rdma *xprt, |
110 | struct xdr_buf *xdr, | |
111 | u32 xdr_off, size_t len, int dir) | |
112 | { | |
113 | struct page *page; | |
114 | dma_addr_t dma_addr; | |
115 | if (xdr_off < xdr->head[0].iov_len) { | |
116 | /* This offset is in the head */ | |
117 | xdr_off += (unsigned long)xdr->head[0].iov_base & ~PAGE_MASK; | |
118 | page = virt_to_page(xdr->head[0].iov_base); | |
119 | } else { | |
120 | xdr_off -= xdr->head[0].iov_len; | |
121 | if (xdr_off < xdr->page_len) { | |
122 | /* This offset is in the page list */ | |
3cbe01a9 | 123 | xdr_off += xdr->page_base; |
b432e6b3 TT |
124 | page = xdr->pages[xdr_off >> PAGE_SHIFT]; |
125 | xdr_off &= ~PAGE_MASK; | |
126 | } else { | |
127 | /* This offset is in the tail */ | |
128 | xdr_off -= xdr->page_len; | |
129 | xdr_off += (unsigned long) | |
130 | xdr->tail[0].iov_base & ~PAGE_MASK; | |
131 | page = virt_to_page(xdr->tail[0].iov_base); | |
132 | } | |
133 | } | |
134 | dma_addr = ib_dma_map_page(xprt->sc_cm_id->device, page, xdr_off, | |
135 | min_t(size_t, PAGE_SIZE, len), dir); | |
136 | return dma_addr; | |
137 | } | |
138 | ||
10dc4512 CL |
139 | /* Returns the address of the first read chunk or <nul> if no read chunk |
140 | * is present | |
141 | */ | |
142 | struct rpcrdma_read_chunk * | |
143 | svc_rdma_get_read_chunk(struct rpcrdma_msg *rmsgp) | |
144 | { | |
145 | struct rpcrdma_read_chunk *ch = | |
146 | (struct rpcrdma_read_chunk *)&rmsgp->rm_body.rm_chunks[0]; | |
147 | ||
148 | if (ch->rc_discrim == xdr_zero) | |
149 | return NULL; | |
150 | return ch; | |
151 | } | |
152 | ||
153 | /* Returns the address of the first read write array element or <nul> | |
154 | * if no write array list is present | |
155 | */ | |
156 | static struct rpcrdma_write_array * | |
157 | svc_rdma_get_write_array(struct rpcrdma_msg *rmsgp) | |
158 | { | |
159 | if (rmsgp->rm_body.rm_chunks[0] != xdr_zero || | |
160 | rmsgp->rm_body.rm_chunks[1] == xdr_zero) | |
161 | return NULL; | |
162 | return (struct rpcrdma_write_array *)&rmsgp->rm_body.rm_chunks[1]; | |
163 | } | |
164 | ||
165 | /* Returns the address of the first reply array element or <nul> if no | |
166 | * reply array is present | |
167 | */ | |
168 | static struct rpcrdma_write_array * | |
169 | svc_rdma_get_reply_array(struct rpcrdma_msg *rmsgp) | |
170 | { | |
171 | struct rpcrdma_read_chunk *rch; | |
172 | struct rpcrdma_write_array *wr_ary; | |
173 | struct rpcrdma_write_array *rp_ary; | |
174 | ||
175 | /* XXX: Need to fix when reply chunk may occur with read list | |
176 | * and/or write list. | |
177 | */ | |
178 | if (rmsgp->rm_body.rm_chunks[0] != xdr_zero || | |
179 | rmsgp->rm_body.rm_chunks[1] != xdr_zero) | |
180 | return NULL; | |
181 | ||
182 | rch = svc_rdma_get_read_chunk(rmsgp); | |
183 | if (rch) { | |
184 | while (rch->rc_discrim != xdr_zero) | |
185 | rch++; | |
186 | ||
187 | /* The reply chunk follows an empty write array located | |
188 | * at 'rc_position' here. The reply array is at rc_target. | |
189 | */ | |
190 | rp_ary = (struct rpcrdma_write_array *)&rch->rc_target; | |
191 | goto found_it; | |
192 | } | |
193 | ||
194 | wr_ary = svc_rdma_get_write_array(rmsgp); | |
195 | if (wr_ary) { | |
196 | int chunk = be32_to_cpu(wr_ary->wc_nchunks); | |
197 | ||
198 | rp_ary = (struct rpcrdma_write_array *) | |
199 | &wr_ary->wc_array[chunk].wc_target.rs_length; | |
200 | goto found_it; | |
201 | } | |
202 | ||
203 | /* No read list, no write list */ | |
204 | rp_ary = (struct rpcrdma_write_array *)&rmsgp->rm_body.rm_chunks[2]; | |
205 | ||
206 | found_it: | |
207 | if (rp_ary->wc_discrim == xdr_zero) | |
208 | return NULL; | |
209 | return rp_ary; | |
210 | } | |
211 | ||
c06b540a TT |
212 | /* Assumptions: |
213 | * - The specified write_len can be represented in sc_max_sge * PAGE_SIZE | |
214 | */ | |
215 | static int send_write(struct svcxprt_rdma *xprt, struct svc_rqst *rqstp, | |
216 | u32 rmr, u64 to, | |
217 | u32 xdr_off, int write_len, | |
34d16e42 | 218 | struct svc_rdma_req_map *vec) |
c06b540a | 219 | { |
e622f2f4 | 220 | struct ib_rdma_wr write_wr; |
c06b540a TT |
221 | struct ib_sge *sge; |
222 | int xdr_sge_no; | |
223 | int sge_no; | |
224 | int sge_bytes; | |
225 | int sge_off; | |
226 | int bc; | |
227 | struct svc_rdma_op_ctxt *ctxt; | |
c06b540a | 228 | |
3fe04ee9 CL |
229 | if (vec->count > RPCSVC_MAXPAGES) { |
230 | pr_err("svcrdma: Too many pages (%lu)\n", vec->count); | |
231 | return -EIO; | |
232 | } | |
233 | ||
c06b540a | 234 | dprintk("svcrdma: RDMA_WRITE rmr=%x, to=%llx, xdr_off=%d, " |
34d16e42 | 235 | "write_len=%d, vec->sge=%p, vec->count=%lu\n", |
bb50c801 | 236 | rmr, (unsigned long long)to, xdr_off, |
34d16e42 | 237 | write_len, vec->sge, vec->count); |
c06b540a TT |
238 | |
239 | ctxt = svc_rdma_get_context(xprt); | |
34d16e42 TT |
240 | ctxt->direction = DMA_TO_DEVICE; |
241 | sge = ctxt->sge; | |
c06b540a TT |
242 | |
243 | /* Find the SGE associated with xdr_off */ | |
34d16e42 | 244 | for (bc = xdr_off, xdr_sge_no = 1; bc && xdr_sge_no < vec->count; |
c06b540a | 245 | xdr_sge_no++) { |
34d16e42 | 246 | if (vec->sge[xdr_sge_no].iov_len > bc) |
c06b540a | 247 | break; |
34d16e42 | 248 | bc -= vec->sge[xdr_sge_no].iov_len; |
c06b540a TT |
249 | } |
250 | ||
251 | sge_off = bc; | |
252 | bc = write_len; | |
253 | sge_no = 0; | |
254 | ||
255 | /* Copy the remaining SGE */ | |
afd566ea TT |
256 | while (bc != 0) { |
257 | sge_bytes = min_t(size_t, | |
258 | bc, vec->sge[xdr_sge_no].iov_len-sge_off); | |
c06b540a | 259 | sge[sge_no].length = sge_bytes; |
0bf48289 SW |
260 | sge[sge_no].addr = |
261 | dma_map_xdr(xprt, &rqstp->rq_res, xdr_off, | |
262 | sge_bytes, DMA_TO_DEVICE); | |
263 | xdr_off += sge_bytes; | |
264 | if (ib_dma_mapping_error(xprt->sc_cm_id->device, | |
265 | sge[sge_no].addr)) | |
266 | goto err; | |
267 | atomic_inc(&xprt->sc_dma_used); | |
268 | sge[sge_no].lkey = xprt->sc_dma_lkey; | |
afd566ea | 269 | ctxt->count++; |
c06b540a TT |
270 | sge_off = 0; |
271 | sge_no++; | |
272 | xdr_sge_no++; | |
3fe04ee9 CL |
273 | if (xdr_sge_no > vec->count) { |
274 | pr_err("svcrdma: Too many sges (%d)\n", xdr_sge_no); | |
275 | goto err; | |
276 | } | |
c06b540a | 277 | bc -= sge_bytes; |
25594290 SW |
278 | if (sge_no == xprt->sc_max_sge) |
279 | break; | |
c06b540a TT |
280 | } |
281 | ||
c06b540a TT |
282 | /* Prepare WRITE WR */ |
283 | memset(&write_wr, 0, sizeof write_wr); | |
284 | ctxt->wr_op = IB_WR_RDMA_WRITE; | |
e622f2f4 CH |
285 | write_wr.wr.wr_id = (unsigned long)ctxt; |
286 | write_wr.wr.sg_list = &sge[0]; | |
287 | write_wr.wr.num_sge = sge_no; | |
288 | write_wr.wr.opcode = IB_WR_RDMA_WRITE; | |
289 | write_wr.wr.send_flags = IB_SEND_SIGNALED; | |
290 | write_wr.rkey = rmr; | |
291 | write_wr.remote_addr = to; | |
c06b540a TT |
292 | |
293 | /* Post It */ | |
294 | atomic_inc(&rdma_stat_write); | |
e622f2f4 | 295 | if (svc_rdma_send(xprt, &write_wr.wr)) |
34d16e42 | 296 | goto err; |
25594290 | 297 | return write_len - bc; |
34d16e42 | 298 | err: |
4a84386f | 299 | svc_rdma_unmap_dma(ctxt); |
34d16e42 TT |
300 | svc_rdma_put_context(ctxt, 0); |
301 | /* Fatal error, close transport */ | |
302 | return -EIO; | |
c06b540a TT |
303 | } |
304 | ||
305 | static int send_write_chunks(struct svcxprt_rdma *xprt, | |
306 | struct rpcrdma_msg *rdma_argp, | |
307 | struct rpcrdma_msg *rdma_resp, | |
308 | struct svc_rqst *rqstp, | |
34d16e42 | 309 | struct svc_rdma_req_map *vec) |
c06b540a TT |
310 | { |
311 | u32 xfer_len = rqstp->rq_res.page_len + rqstp->rq_res.tail[0].iov_len; | |
312 | int write_len; | |
c06b540a TT |
313 | u32 xdr_off; |
314 | int chunk_off; | |
315 | int chunk_no; | |
70747c25 | 316 | int nchunks; |
c06b540a TT |
317 | struct rpcrdma_write_array *arg_ary; |
318 | struct rpcrdma_write_array *res_ary; | |
319 | int ret; | |
320 | ||
321 | arg_ary = svc_rdma_get_write_array(rdma_argp); | |
322 | if (!arg_ary) | |
323 | return 0; | |
324 | res_ary = (struct rpcrdma_write_array *) | |
325 | &rdma_resp->rm_body.rm_chunks[1]; | |
326 | ||
c06b540a | 327 | /* Write chunks start at the pagelist */ |
70747c25 | 328 | nchunks = be32_to_cpu(arg_ary->wc_nchunks); |
c06b540a | 329 | for (xdr_off = rqstp->rq_res.head[0].iov_len, chunk_no = 0; |
70747c25 | 330 | xfer_len && chunk_no < nchunks; |
c06b540a TT |
331 | chunk_no++) { |
332 | struct rpcrdma_segment *arg_ch; | |
333 | u64 rs_offset; | |
334 | ||
335 | arg_ch = &arg_ary->wc_array[chunk_no].wc_target; | |
70747c25 | 336 | write_len = min(xfer_len, be32_to_cpu(arg_ch->rs_length)); |
c06b540a TT |
337 | |
338 | /* Prepare the response chunk given the length actually | |
339 | * written */ | |
cec56c8f | 340 | xdr_decode_hyper((__be32 *)&arg_ch->rs_offset, &rs_offset); |
c06b540a | 341 | svc_rdma_xdr_encode_array_chunk(res_ary, chunk_no, |
cec56c8f TT |
342 | arg_ch->rs_handle, |
343 | arg_ch->rs_offset, | |
344 | write_len); | |
c06b540a TT |
345 | chunk_off = 0; |
346 | while (write_len) { | |
c06b540a | 347 | ret = send_write(xprt, rqstp, |
70747c25 | 348 | be32_to_cpu(arg_ch->rs_handle), |
c06b540a TT |
349 | rs_offset + chunk_off, |
350 | xdr_off, | |
25594290 | 351 | write_len, |
34d16e42 | 352 | vec); |
25594290 | 353 | if (ret <= 0) { |
c06b540a TT |
354 | dprintk("svcrdma: RDMA_WRITE failed, ret=%d\n", |
355 | ret); | |
356 | return -EIO; | |
357 | } | |
25594290 SW |
358 | chunk_off += ret; |
359 | xdr_off += ret; | |
360 | xfer_len -= ret; | |
361 | write_len -= ret; | |
c06b540a TT |
362 | } |
363 | } | |
364 | /* Update the req with the number of chunks actually used */ | |
365 | svc_rdma_xdr_encode_write_list(rdma_resp, chunk_no); | |
366 | ||
367 | return rqstp->rq_res.page_len + rqstp->rq_res.tail[0].iov_len; | |
368 | } | |
369 | ||
370 | static int send_reply_chunks(struct svcxprt_rdma *xprt, | |
371 | struct rpcrdma_msg *rdma_argp, | |
372 | struct rpcrdma_msg *rdma_resp, | |
373 | struct svc_rqst *rqstp, | |
34d16e42 | 374 | struct svc_rdma_req_map *vec) |
c06b540a TT |
375 | { |
376 | u32 xfer_len = rqstp->rq_res.len; | |
377 | int write_len; | |
c06b540a TT |
378 | u32 xdr_off; |
379 | int chunk_no; | |
380 | int chunk_off; | |
cec56c8f | 381 | int nchunks; |
c06b540a TT |
382 | struct rpcrdma_segment *ch; |
383 | struct rpcrdma_write_array *arg_ary; | |
384 | struct rpcrdma_write_array *res_ary; | |
385 | int ret; | |
386 | ||
387 | arg_ary = svc_rdma_get_reply_array(rdma_argp); | |
388 | if (!arg_ary) | |
389 | return 0; | |
390 | /* XXX: need to fix when reply lists occur with read-list and or | |
391 | * write-list */ | |
392 | res_ary = (struct rpcrdma_write_array *) | |
393 | &rdma_resp->rm_body.rm_chunks[2]; | |
394 | ||
c06b540a | 395 | /* xdr offset starts at RPC message */ |
70747c25 | 396 | nchunks = be32_to_cpu(arg_ary->wc_nchunks); |
c06b540a | 397 | for (xdr_off = 0, chunk_no = 0; |
cec56c8f | 398 | xfer_len && chunk_no < nchunks; |
c06b540a TT |
399 | chunk_no++) { |
400 | u64 rs_offset; | |
401 | ch = &arg_ary->wc_array[chunk_no].wc_target; | |
70747c25 | 402 | write_len = min(xfer_len, be32_to_cpu(ch->rs_length)); |
c06b540a | 403 | |
c06b540a TT |
404 | /* Prepare the reply chunk given the length actually |
405 | * written */ | |
cec56c8f | 406 | xdr_decode_hyper((__be32 *)&ch->rs_offset, &rs_offset); |
c06b540a | 407 | svc_rdma_xdr_encode_array_chunk(res_ary, chunk_no, |
cec56c8f TT |
408 | ch->rs_handle, ch->rs_offset, |
409 | write_len); | |
c06b540a TT |
410 | chunk_off = 0; |
411 | while (write_len) { | |
c06b540a | 412 | ret = send_write(xprt, rqstp, |
70747c25 | 413 | be32_to_cpu(ch->rs_handle), |
c06b540a TT |
414 | rs_offset + chunk_off, |
415 | xdr_off, | |
25594290 | 416 | write_len, |
34d16e42 | 417 | vec); |
25594290 | 418 | if (ret <= 0) { |
c06b540a TT |
419 | dprintk("svcrdma: RDMA_WRITE failed, ret=%d\n", |
420 | ret); | |
421 | return -EIO; | |
422 | } | |
25594290 SW |
423 | chunk_off += ret; |
424 | xdr_off += ret; | |
425 | xfer_len -= ret; | |
426 | write_len -= ret; | |
c06b540a TT |
427 | } |
428 | } | |
429 | /* Update the req with the number of chunks actually used */ | |
430 | svc_rdma_xdr_encode_reply_array(res_ary, chunk_no); | |
431 | ||
432 | return rqstp->rq_res.len; | |
433 | } | |
434 | ||
435 | /* This function prepares the portion of the RPCRDMA message to be | |
436 | * sent in the RDMA_SEND. This function is called after data sent via | |
437 | * RDMA has already been transmitted. There are three cases: | |
438 | * - The RPCRDMA header, RPC header, and payload are all sent in a | |
439 | * single RDMA_SEND. This is the "inline" case. | |
440 | * - The RPCRDMA header and some portion of the RPC header and data | |
441 | * are sent via this RDMA_SEND and another portion of the data is | |
442 | * sent via RDMA. | |
443 | * - The RPCRDMA header [NOMSG] is sent in this RDMA_SEND and the RPC | |
444 | * header and data are all transmitted via RDMA. | |
445 | * In all three cases, this function prepares the RPCRDMA header in | |
446 | * sge[0], the 'type' parameter indicates the type to place in the | |
447 | * RPCRDMA header, and the 'byte_count' field indicates how much of | |
b432e6b3 TT |
448 | * the XDR to include in this RDMA_SEND. NB: The offset of the payload |
449 | * to send is zero in the XDR. | |
c06b540a TT |
450 | */ |
451 | static int send_reply(struct svcxprt_rdma *rdma, | |
452 | struct svc_rqst *rqstp, | |
453 | struct page *page, | |
454 | struct rpcrdma_msg *rdma_resp, | |
455 | struct svc_rdma_op_ctxt *ctxt, | |
34d16e42 | 456 | struct svc_rdma_req_map *vec, |
c06b540a TT |
457 | int byte_count) |
458 | { | |
459 | struct ib_send_wr send_wr; | |
9d11b51c | 460 | u32 xdr_off; |
c06b540a TT |
461 | int sge_no; |
462 | int sge_bytes; | |
463 | int page_no; | |
afc59400 | 464 | int pages; |
c06b540a TT |
465 | int ret; |
466 | ||
0e7f011a TT |
467 | /* Post a recv buffer to handle another request. */ |
468 | ret = svc_rdma_post_recv(rdma); | |
469 | if (ret) { | |
470 | printk(KERN_INFO | |
471 | "svcrdma: could not post a receive buffer, err=%d." | |
472 | "Closing transport %p.\n", ret, rdma); | |
473 | set_bit(XPT_CLOSE, &rdma->sc_xprt.xpt_flags); | |
5ac461a6 TT |
474 | svc_rdma_put_context(ctxt, 0); |
475 | return -ENOTCONN; | |
0e7f011a TT |
476 | } |
477 | ||
c06b540a TT |
478 | /* Prepare the context */ |
479 | ctxt->pages[0] = page; | |
480 | ctxt->count = 1; | |
481 | ||
482 | /* Prepare the SGE for the RPCRDMA Header */ | |
98779be8 SW |
483 | ctxt->sge[0].lkey = rdma->sc_dma_lkey; |
484 | ctxt->sge[0].length = svc_rdma_xdr_get_reply_hdr_len(rdma_resp); | |
c06b540a | 485 | ctxt->sge[0].addr = |
b432e6b3 TT |
486 | ib_dma_map_page(rdma->sc_cm_id->device, page, 0, |
487 | ctxt->sge[0].length, DMA_TO_DEVICE); | |
afd566ea TT |
488 | if (ib_dma_mapping_error(rdma->sc_cm_id->device, ctxt->sge[0].addr)) |
489 | goto err; | |
490 | atomic_inc(&rdma->sc_dma_used); | |
491 | ||
c06b540a | 492 | ctxt->direction = DMA_TO_DEVICE; |
afd566ea | 493 | |
b432e6b3 | 494 | /* Map the payload indicated by 'byte_count' */ |
9d11b51c | 495 | xdr_off = 0; |
34d16e42 TT |
496 | for (sge_no = 1; byte_count && sge_no < vec->count; sge_no++) { |
497 | sge_bytes = min_t(size_t, vec->sge[sge_no].iov_len, byte_count); | |
c06b540a | 498 | byte_count -= sge_bytes; |
0bf48289 SW |
499 | ctxt->sge[sge_no].addr = |
500 | dma_map_xdr(rdma, &rqstp->rq_res, xdr_off, | |
501 | sge_bytes, DMA_TO_DEVICE); | |
502 | xdr_off += sge_bytes; | |
503 | if (ib_dma_mapping_error(rdma->sc_cm_id->device, | |
504 | ctxt->sge[sge_no].addr)) | |
505 | goto err; | |
506 | atomic_inc(&rdma->sc_dma_used); | |
507 | ctxt->sge[sge_no].lkey = rdma->sc_dma_lkey; | |
34d16e42 | 508 | ctxt->sge[sge_no].length = sge_bytes; |
c06b540a | 509 | } |
3fe04ee9 CL |
510 | if (byte_count != 0) { |
511 | pr_err("svcrdma: Could not map %d bytes\n", byte_count); | |
512 | goto err; | |
513 | } | |
c06b540a TT |
514 | |
515 | /* Save all respages in the ctxt and remove them from the | |
516 | * respages array. They are our pages until the I/O | |
517 | * completes. | |
518 | */ | |
afc59400 BF |
519 | pages = rqstp->rq_next_page - rqstp->rq_respages; |
520 | for (page_no = 0; page_no < pages; page_no++) { | |
c06b540a TT |
521 | ctxt->pages[page_no+1] = rqstp->rq_respages[page_no]; |
522 | ctxt->count++; | |
523 | rqstp->rq_respages[page_no] = NULL; | |
afd566ea TT |
524 | /* |
525 | * If there are more pages than SGE, terminate SGE | |
526 | * list so that svc_rdma_unmap_dma doesn't attempt to | |
527 | * unmap garbage. | |
528 | */ | |
34d16e42 TT |
529 | if (page_no+1 >= sge_no) |
530 | ctxt->sge[page_no+1].length = 0; | |
c06b540a | 531 | } |
7e4359e2 | 532 | rqstp->rq_next_page = rqstp->rq_respages + 1; |
0bf48289 | 533 | |
9d11b51c CL |
534 | /* The loop above bumps sc_dma_used for each sge. The |
535 | * xdr_buf.tail gets a separate sge, but resides in the | |
536 | * same page as xdr_buf.head. Don't count it twice. | |
537 | */ | |
538 | if (sge_no > ctxt->count) | |
539 | atomic_dec(&rdma->sc_dma_used); | |
540 | ||
3fe04ee9 CL |
541 | if (sge_no > rdma->sc_max_sge) { |
542 | pr_err("svcrdma: Too many sges (%d)\n", sge_no); | |
543 | goto err; | |
544 | } | |
c06b540a TT |
545 | memset(&send_wr, 0, sizeof send_wr); |
546 | ctxt->wr_op = IB_WR_SEND; | |
547 | send_wr.wr_id = (unsigned long)ctxt; | |
548 | send_wr.sg_list = ctxt->sge; | |
549 | send_wr.num_sge = sge_no; | |
550 | send_wr.opcode = IB_WR_SEND; | |
551 | send_wr.send_flags = IB_SEND_SIGNALED; | |
552 | ||
553 | ret = svc_rdma_send(rdma, &send_wr); | |
554 | if (ret) | |
afd566ea | 555 | goto err; |
c06b540a | 556 | |
afd566ea TT |
557 | return 0; |
558 | ||
559 | err: | |
21515e46 | 560 | svc_rdma_unmap_dma(ctxt); |
afd566ea TT |
561 | svc_rdma_put_context(ctxt, 1); |
562 | return -EIO; | |
c06b540a TT |
563 | } |
564 | ||
565 | void svc_rdma_prep_reply_hdr(struct svc_rqst *rqstp) | |
566 | { | |
567 | } | |
568 | ||
c06b540a TT |
569 | int svc_rdma_sendto(struct svc_rqst *rqstp) |
570 | { | |
571 | struct svc_xprt *xprt = rqstp->rq_xprt; | |
572 | struct svcxprt_rdma *rdma = | |
573 | container_of(xprt, struct svcxprt_rdma, sc_xprt); | |
574 | struct rpcrdma_msg *rdma_argp; | |
575 | struct rpcrdma_msg *rdma_resp; | |
576 | struct rpcrdma_write_array *reply_ary; | |
577 | enum rpcrdma_proc reply_type; | |
578 | int ret; | |
579 | int inline_bytes; | |
c06b540a TT |
580 | struct page *res_page; |
581 | struct svc_rdma_op_ctxt *ctxt; | |
34d16e42 | 582 | struct svc_rdma_req_map *vec; |
c06b540a TT |
583 | |
584 | dprintk("svcrdma: sending response for rqstp=%p\n", rqstp); | |
585 | ||
e5523bd2 CL |
586 | /* Get the RDMA request header. The receive logic always |
587 | * places this at the start of page 0. | |
588 | */ | |
589 | rdma_argp = page_address(rqstp->rq_pages[0]); | |
c06b540a | 590 | |
34d16e42 | 591 | /* Build an req vec for the XDR */ |
c06b540a TT |
592 | ctxt = svc_rdma_get_context(rdma); |
593 | ctxt->direction = DMA_TO_DEVICE; | |
34d16e42 | 594 | vec = svc_rdma_get_req_map(); |
afd566ea TT |
595 | ret = map_xdr(rdma, &rqstp->rq_res, vec); |
596 | if (ret) | |
597 | goto err0; | |
c06b540a TT |
598 | inline_bytes = rqstp->rq_res.len; |
599 | ||
600 | /* Create the RDMA response header */ | |
b7e0b9a9 | 601 | res_page = alloc_page(GFP_KERNEL | __GFP_NOFAIL); |
c06b540a TT |
602 | rdma_resp = page_address(res_page); |
603 | reply_ary = svc_rdma_get_reply_array(rdma_argp); | |
604 | if (reply_ary) | |
605 | reply_type = RDMA_NOMSG; | |
606 | else | |
607 | reply_type = RDMA_MSG; | |
608 | svc_rdma_xdr_encode_reply_header(rdma, rdma_argp, | |
609 | rdma_resp, reply_type); | |
610 | ||
611 | /* Send any write-chunk data and build resp write-list */ | |
612 | ret = send_write_chunks(rdma, rdma_argp, rdma_resp, | |
34d16e42 | 613 | rqstp, vec); |
c06b540a TT |
614 | if (ret < 0) { |
615 | printk(KERN_ERR "svcrdma: failed to send write chunks, rc=%d\n", | |
616 | ret); | |
afd566ea | 617 | goto err1; |
c06b540a TT |
618 | } |
619 | inline_bytes -= ret; | |
620 | ||
621 | /* Send any reply-list data and update resp reply-list */ | |
622 | ret = send_reply_chunks(rdma, rdma_argp, rdma_resp, | |
34d16e42 | 623 | rqstp, vec); |
c06b540a TT |
624 | if (ret < 0) { |
625 | printk(KERN_ERR "svcrdma: failed to send reply chunks, rc=%d\n", | |
626 | ret); | |
afd566ea | 627 | goto err1; |
c06b540a TT |
628 | } |
629 | inline_bytes -= ret; | |
630 | ||
34d16e42 | 631 | ret = send_reply(rdma, rqstp, res_page, rdma_resp, ctxt, vec, |
c06b540a | 632 | inline_bytes); |
34d16e42 | 633 | svc_rdma_put_req_map(vec); |
c06b540a TT |
634 | dprintk("svcrdma: send_reply returns %d\n", ret); |
635 | return ret; | |
afd566ea TT |
636 | |
637 | err1: | |
638 | put_page(res_page); | |
639 | err0: | |
34d16e42 | 640 | svc_rdma_put_req_map(vec); |
c06b540a | 641 | svc_rdma_put_context(ctxt, 0); |
c06b540a TT |
642 | return ret; |
643 | } |