1 // SPDX-License-Identifier: GPL-2.0
2 #include <linux/ceph/ceph_debug.h>
3
4 #include <linux/bvec.h>
5 #include <linux/crc32c.h>
6 #include <linux/net.h>
7 #include <linux/socket.h>
8 #include <net/sock.h>
9
10 #include <linux/ceph/ceph_features.h>
11 #include <linux/ceph/decode.h>
12 #include <linux/ceph/libceph.h>
13 #include <linux/ceph/messenger.h>
14
15 /* static tag bytes (protocol control messages) */
16 static char tag_msg = CEPH_MSGR_TAG_MSG;
17 static char tag_ack = CEPH_MSGR_TAG_ACK;
18 static char tag_keepalive = CEPH_MSGR_TAG_KEEPALIVE;
19 static char tag_keepalive2 = CEPH_MSGR_TAG_KEEPALIVE2;
20
21 /*
22 * If @buf is NULL, discard up to @len bytes.
23 */
ceph_tcp_recvmsg(struct socket * sock,void * buf,size_t len)24 static int ceph_tcp_recvmsg(struct socket *sock, void *buf, size_t len)
25 {
26 struct kvec iov = {buf, len};
27 struct msghdr msg = { .msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL };
28 int r;
29
30 if (!buf)
31 msg.msg_flags |= MSG_TRUNC;
32
33 iov_iter_kvec(&msg.msg_iter, READ, &iov, 1, len);
34 r = sock_recvmsg(sock, &msg, msg.msg_flags);
35 if (r == -EAGAIN)
36 r = 0;
37 return r;
38 }
39
ceph_tcp_recvpage(struct socket * sock,struct page * page,int page_offset,size_t length)40 static int ceph_tcp_recvpage(struct socket *sock, struct page *page,
41 int page_offset, size_t length)
42 {
43 struct bio_vec bvec = {
44 .bv_page = page,
45 .bv_offset = page_offset,
46 .bv_len = length
47 };
48 struct msghdr msg = { .msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL };
49 int r;
50
51 BUG_ON(page_offset + length > PAGE_SIZE);
52 iov_iter_bvec(&msg.msg_iter, READ, &bvec, 1, length);
53 r = sock_recvmsg(sock, &msg, msg.msg_flags);
54 if (r == -EAGAIN)
55 r = 0;
56 return r;
57 }
58
59 /*
60 * write something. @more is true if caller will be sending more data
61 * shortly.
62 */
ceph_tcp_sendmsg(struct socket * sock,struct kvec * iov,size_t kvlen,size_t len,bool more)63 static int ceph_tcp_sendmsg(struct socket *sock, struct kvec *iov,
64 size_t kvlen, size_t len, bool more)
65 {
66 struct msghdr msg = { .msg_flags = MSG_DONTWAIT | MSG_NOSIGNAL };
67 int r;
68
69 if (more)
70 msg.msg_flags |= MSG_MORE;
71 else
72 msg.msg_flags |= MSG_EOR; /* superfluous, but what the hell */
73
74 r = kernel_sendmsg(sock, &msg, iov, kvlen, len);
75 if (r == -EAGAIN)
76 r = 0;
77 return r;
78 }
79
80 /*
81 * @more: either or both of MSG_MORE and MSG_SENDPAGE_NOTLAST
82 */
ceph_tcp_sendpage(struct socket * sock,struct page * page,int offset,size_t size,int more)83 static int ceph_tcp_sendpage(struct socket *sock, struct page *page,
84 int offset, size_t size, int more)
85 {
86 ssize_t (*sendpage)(struct socket *sock, struct page *page,
87 int offset, size_t size, int flags);
88 int flags = MSG_DONTWAIT | MSG_NOSIGNAL | more;
89 int ret;
90
91 /*
92 * sendpage cannot properly handle pages with page_count == 0,
93 * we need to fall back to sendmsg if that's the case.
94 *
95 * Same goes for slab pages: skb_can_coalesce() allows
96 * coalescing neighboring slab objects into a single frag which
97 * triggers one of hardened usercopy checks.
98 */
99 if (sendpage_ok(page))
100 sendpage = sock->ops->sendpage;
101 else
102 sendpage = sock_no_sendpage;
103
104 ret = sendpage(sock, page, offset, size, flags);
105 if (ret == -EAGAIN)
106 ret = 0;
107
108 return ret;
109 }
110
con_out_kvec_reset(struct ceph_connection * con)111 static void con_out_kvec_reset(struct ceph_connection *con)
112 {
113 BUG_ON(con->v1.out_skip);
114
115 con->v1.out_kvec_left = 0;
116 con->v1.out_kvec_bytes = 0;
117 con->v1.out_kvec_cur = &con->v1.out_kvec[0];
118 }
119
con_out_kvec_add(struct ceph_connection * con,size_t size,void * data)120 static void con_out_kvec_add(struct ceph_connection *con,
121 size_t size, void *data)
122 {
123 int index = con->v1.out_kvec_left;
124
125 BUG_ON(con->v1.out_skip);
126 BUG_ON(index >= ARRAY_SIZE(con->v1.out_kvec));
127
128 con->v1.out_kvec[index].iov_len = size;
129 con->v1.out_kvec[index].iov_base = data;
130 con->v1.out_kvec_left++;
131 con->v1.out_kvec_bytes += size;
132 }
133
134 /*
135 * Chop off a kvec from the end. Return residual number of bytes for
136 * that kvec, i.e. how many bytes would have been written if the kvec
137 * hadn't been nuked.
138 */
con_out_kvec_skip(struct ceph_connection * con)139 static int con_out_kvec_skip(struct ceph_connection *con)
140 {
141 int skip = 0;
142
143 if (con->v1.out_kvec_bytes > 0) {
144 skip = con->v1.out_kvec_cur[con->v1.out_kvec_left - 1].iov_len;
145 BUG_ON(con->v1.out_kvec_bytes < skip);
146 BUG_ON(!con->v1.out_kvec_left);
147 con->v1.out_kvec_bytes -= skip;
148 con->v1.out_kvec_left--;
149 }
150
151 return skip;
152 }
153
sizeof_footer(struct ceph_connection * con)154 static size_t sizeof_footer(struct ceph_connection *con)
155 {
156 return (con->peer_features & CEPH_FEATURE_MSG_AUTH) ?
157 sizeof(struct ceph_msg_footer) :
158 sizeof(struct ceph_msg_footer_old);
159 }
160
prepare_message_data(struct ceph_msg * msg,u32 data_len)161 static void prepare_message_data(struct ceph_msg *msg, u32 data_len)
162 {
163 /* Initialize data cursor */
164
165 ceph_msg_data_cursor_init(&msg->cursor, msg, data_len);
166 }
167
168 /*
169 * Prepare footer for currently outgoing message, and finish things
170 * off. Assumes out_kvec* are already valid.. we just add on to the end.
171 */
prepare_write_message_footer(struct ceph_connection * con)172 static void prepare_write_message_footer(struct ceph_connection *con)
173 {
174 struct ceph_msg *m = con->out_msg;
175
176 m->footer.flags |= CEPH_MSG_FOOTER_COMPLETE;
177
178 dout("prepare_write_message_footer %p\n", con);
179 con_out_kvec_add(con, sizeof_footer(con), &m->footer);
180 if (con->peer_features & CEPH_FEATURE_MSG_AUTH) {
181 if (con->ops->sign_message)
182 con->ops->sign_message(m);
183 else
184 m->footer.sig = 0;
185 } else {
186 m->old_footer.flags = m->footer.flags;
187 }
188 con->v1.out_more = m->more_to_follow;
189 con->v1.out_msg_done = true;
190 }
191
192 /*
193 * Prepare headers for the next outgoing message.
194 */
prepare_write_message(struct ceph_connection * con)195 static void prepare_write_message(struct ceph_connection *con)
196 {
197 struct ceph_msg *m;
198 u32 crc;
199
200 con_out_kvec_reset(con);
201 con->v1.out_msg_done = false;
202
203 /* Sneak an ack in there first? If we can get it into the same
204 * TCP packet that's a good thing. */
205 if (con->in_seq > con->in_seq_acked) {
206 con->in_seq_acked = con->in_seq;
207 con_out_kvec_add(con, sizeof (tag_ack), &tag_ack);
208 con->v1.out_temp_ack = cpu_to_le64(con->in_seq_acked);
209 con_out_kvec_add(con, sizeof(con->v1.out_temp_ack),
210 &con->v1.out_temp_ack);
211 }
212
213 ceph_con_get_out_msg(con);
214 m = con->out_msg;
215
216 dout("prepare_write_message %p seq %lld type %d len %d+%d+%zd\n",
217 m, con->out_seq, le16_to_cpu(m->hdr.type),
218 le32_to_cpu(m->hdr.front_len), le32_to_cpu(m->hdr.middle_len),
219 m->data_length);
220 WARN_ON(m->front.iov_len != le32_to_cpu(m->hdr.front_len));
221 WARN_ON(m->data_length != le32_to_cpu(m->hdr.data_len));
222
223 /* tag + hdr + front + middle */
224 con_out_kvec_add(con, sizeof (tag_msg), &tag_msg);
225 con_out_kvec_add(con, sizeof(con->v1.out_hdr), &con->v1.out_hdr);
226 con_out_kvec_add(con, m->front.iov_len, m->front.iov_base);
227
228 if (m->middle)
229 con_out_kvec_add(con, m->middle->vec.iov_len,
230 m->middle->vec.iov_base);
231
232 /* fill in hdr crc and finalize hdr */
233 crc = crc32c(0, &m->hdr, offsetof(struct ceph_msg_header, crc));
234 con->out_msg->hdr.crc = cpu_to_le32(crc);
235 memcpy(&con->v1.out_hdr, &con->out_msg->hdr, sizeof(con->v1.out_hdr));
236
237 /* fill in front and middle crc, footer */
238 crc = crc32c(0, m->front.iov_base, m->front.iov_len);
239 con->out_msg->footer.front_crc = cpu_to_le32(crc);
240 if (m->middle) {
241 crc = crc32c(0, m->middle->vec.iov_base,
242 m->middle->vec.iov_len);
243 con->out_msg->footer.middle_crc = cpu_to_le32(crc);
244 } else
245 con->out_msg->footer.middle_crc = 0;
246 dout("%s front_crc %u middle_crc %u\n", __func__,
247 le32_to_cpu(con->out_msg->footer.front_crc),
248 le32_to_cpu(con->out_msg->footer.middle_crc));
249 con->out_msg->footer.flags = 0;
250
251 /* is there a data payload? */
252 con->out_msg->footer.data_crc = 0;
253 if (m->data_length) {
254 prepare_message_data(con->out_msg, m->data_length);
255 con->v1.out_more = 1; /* data + footer will follow */
256 } else {
257 /* no, queue up footer too and be done */
258 prepare_write_message_footer(con);
259 }
260
261 ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
262 }
263
264 /*
265 * Prepare an ack.
266 */
prepare_write_ack(struct ceph_connection * con)267 static void prepare_write_ack(struct ceph_connection *con)
268 {
269 dout("prepare_write_ack %p %llu -> %llu\n", con,
270 con->in_seq_acked, con->in_seq);
271 con->in_seq_acked = con->in_seq;
272
273 con_out_kvec_reset(con);
274
275 con_out_kvec_add(con, sizeof (tag_ack), &tag_ack);
276
277 con->v1.out_temp_ack = cpu_to_le64(con->in_seq_acked);
278 con_out_kvec_add(con, sizeof(con->v1.out_temp_ack),
279 &con->v1.out_temp_ack);
280
281 con->v1.out_more = 1; /* more will follow.. eventually.. */
282 ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
283 }
284
285 /*
286 * Prepare to share the seq during handshake
287 */
prepare_write_seq(struct ceph_connection * con)288 static void prepare_write_seq(struct ceph_connection *con)
289 {
290 dout("prepare_write_seq %p %llu -> %llu\n", con,
291 con->in_seq_acked, con->in_seq);
292 con->in_seq_acked = con->in_seq;
293
294 con_out_kvec_reset(con);
295
296 con->v1.out_temp_ack = cpu_to_le64(con->in_seq_acked);
297 con_out_kvec_add(con, sizeof(con->v1.out_temp_ack),
298 &con->v1.out_temp_ack);
299
300 ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
301 }
302
303 /*
304 * Prepare to write keepalive byte.
305 */
prepare_write_keepalive(struct ceph_connection * con)306 static void prepare_write_keepalive(struct ceph_connection *con)
307 {
308 dout("prepare_write_keepalive %p\n", con);
309 con_out_kvec_reset(con);
310 if (con->peer_features & CEPH_FEATURE_MSGR_KEEPALIVE2) {
311 struct timespec64 now;
312
313 ktime_get_real_ts64(&now);
314 con_out_kvec_add(con, sizeof(tag_keepalive2), &tag_keepalive2);
315 ceph_encode_timespec64(&con->v1.out_temp_keepalive2, &now);
316 con_out_kvec_add(con, sizeof(con->v1.out_temp_keepalive2),
317 &con->v1.out_temp_keepalive2);
318 } else {
319 con_out_kvec_add(con, sizeof(tag_keepalive), &tag_keepalive);
320 }
321 ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
322 }
323
324 /*
325 * Connection negotiation.
326 */
327
get_connect_authorizer(struct ceph_connection * con)328 static int get_connect_authorizer(struct ceph_connection *con)
329 {
330 struct ceph_auth_handshake *auth;
331 int auth_proto;
332
333 if (!con->ops->get_authorizer) {
334 con->v1.auth = NULL;
335 con->v1.out_connect.authorizer_protocol = CEPH_AUTH_UNKNOWN;
336 con->v1.out_connect.authorizer_len = 0;
337 return 0;
338 }
339
340 auth = con->ops->get_authorizer(con, &auth_proto, con->v1.auth_retry);
341 if (IS_ERR(auth))
342 return PTR_ERR(auth);
343
344 con->v1.auth = auth;
345 con->v1.out_connect.authorizer_protocol = cpu_to_le32(auth_proto);
346 con->v1.out_connect.authorizer_len =
347 cpu_to_le32(auth->authorizer_buf_len);
348 return 0;
349 }
350
351 /*
352 * We connected to a peer and are saying hello.
353 */
prepare_write_banner(struct ceph_connection * con)354 static void prepare_write_banner(struct ceph_connection *con)
355 {
356 con_out_kvec_add(con, strlen(CEPH_BANNER), CEPH_BANNER);
357 con_out_kvec_add(con, sizeof (con->msgr->my_enc_addr),
358 &con->msgr->my_enc_addr);
359
360 con->v1.out_more = 0;
361 ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
362 }
363
__prepare_write_connect(struct ceph_connection * con)364 static void __prepare_write_connect(struct ceph_connection *con)
365 {
366 con_out_kvec_add(con, sizeof(con->v1.out_connect),
367 &con->v1.out_connect);
368 if (con->v1.auth)
369 con_out_kvec_add(con, con->v1.auth->authorizer_buf_len,
370 con->v1.auth->authorizer_buf);
371
372 con->v1.out_more = 0;
373 ceph_con_flag_set(con, CEPH_CON_F_WRITE_PENDING);
374 }
375
prepare_write_connect(struct ceph_connection * con)376 static int prepare_write_connect(struct ceph_connection *con)
377 {
378 unsigned int global_seq = ceph_get_global_seq(con->msgr, 0);
379 int proto;
380 int ret;
381
382 switch (con->peer_name.type) {
383 case CEPH_ENTITY_TYPE_MON:
384 proto = CEPH_MONC_PROTOCOL;
385 break;
386 case CEPH_ENTITY_TYPE_OSD:
387 proto = CEPH_OSDC_PROTOCOL;
388 break;
389 case CEPH_ENTITY_TYPE_MDS:
390 proto = CEPH_MDSC_PROTOCOL;
391 break;
392 default:
393 BUG();
394 }
395
396 dout("prepare_write_connect %p cseq=%d gseq=%d proto=%d\n", con,
397 con->v1.connect_seq, global_seq, proto);
398
399 con->v1.out_connect.features =
400 cpu_to_le64(from_msgr(con->msgr)->supported_features);
401 con->v1.out_connect.host_type = cpu_to_le32(CEPH_ENTITY_TYPE_CLIENT);
402 con->v1.out_connect.connect_seq = cpu_to_le32(con->v1.connect_seq);
403 con->v1.out_connect.global_seq = cpu_to_le32(global_seq);
404 con->v1.out_connect.protocol_version = cpu_to_le32(proto);
405 con->v1.out_connect.flags = 0;
406
407 ret = get_connect_authorizer(con);
408 if (ret)
409 return ret;
410
411 __prepare_write_connect(con);
412 return 0;
413 }
414
415 /*
416 * write as much of pending kvecs to the socket as we can.
417 * 1 -> done
418 * 0 -> socket full, but more to do
419 * <0 -> error
420 */
write_partial_kvec(struct ceph_connection * con)421 static int write_partial_kvec(struct ceph_connection *con)
422 {
423 int ret;
424
425 dout("write_partial_kvec %p %d left\n", con, con->v1.out_kvec_bytes);
426 while (con->v1.out_kvec_bytes > 0) {
427 ret = ceph_tcp_sendmsg(con->sock, con->v1.out_kvec_cur,
428 con->v1.out_kvec_left,
429 con->v1.out_kvec_bytes,
430 con->v1.out_more);
431 if (ret <= 0)
432 goto out;
433 con->v1.out_kvec_bytes -= ret;
434 if (!con->v1.out_kvec_bytes)
435 break; /* done */
436
437 /* account for full iov entries consumed */
438 while (ret >= con->v1.out_kvec_cur->iov_len) {
439 BUG_ON(!con->v1.out_kvec_left);
440 ret -= con->v1.out_kvec_cur->iov_len;
441 con->v1.out_kvec_cur++;
442 con->v1.out_kvec_left--;
443 }
444 /* and for a partially-consumed entry */
445 if (ret) {
446 con->v1.out_kvec_cur->iov_len -= ret;
447 con->v1.out_kvec_cur->iov_base += ret;
448 }
449 }
450 con->v1.out_kvec_left = 0;
451 ret = 1;
452 out:
453 dout("write_partial_kvec %p %d left in %d kvecs ret = %d\n", con,
454 con->v1.out_kvec_bytes, con->v1.out_kvec_left, ret);
455 return ret; /* done! */
456 }
457
458 /*
459 * Write as much message data payload as we can. If we finish, queue
460 * up the footer.
461 * 1 -> done, footer is now queued in out_kvec[].
462 * 0 -> socket full, but more to do
463 * <0 -> error
464 */
write_partial_message_data(struct ceph_connection * con)465 static int write_partial_message_data(struct ceph_connection *con)
466 {
467 struct ceph_msg *msg = con->out_msg;
468 struct ceph_msg_data_cursor *cursor = &msg->cursor;
469 bool do_datacrc = !ceph_test_opt(from_msgr(con->msgr), NOCRC);
470 int more = MSG_MORE | MSG_SENDPAGE_NOTLAST;
471 u32 crc;
472
473 dout("%s %p msg %p\n", __func__, con, msg);
474
475 if (!msg->num_data_items)
476 return -EINVAL;
477
478 /*
479 * Iterate through each page that contains data to be
480 * written, and send as much as possible for each.
481 *
482 * If we are calculating the data crc (the default), we will
483 * need to map the page. If we have no pages, they have
484 * been revoked, so use the zero page.
485 */
486 crc = do_datacrc ? le32_to_cpu(msg->footer.data_crc) : 0;
487 while (cursor->total_resid) {
488 struct page *page;
489 size_t page_offset;
490 size_t length;
491 int ret;
492
493 if (!cursor->resid) {
494 ceph_msg_data_advance(cursor, 0);
495 continue;
496 }
497
498 page = ceph_msg_data_next(cursor, &page_offset, &length, NULL);
499 if (length == cursor->total_resid)
500 more = MSG_MORE;
501 ret = ceph_tcp_sendpage(con->sock, page, page_offset, length,
502 more);
503 if (ret <= 0) {
504 if (do_datacrc)
505 msg->footer.data_crc = cpu_to_le32(crc);
506
507 return ret;
508 }
509 if (do_datacrc && cursor->need_crc)
510 crc = ceph_crc32c_page(crc, page, page_offset, length);
511 ceph_msg_data_advance(cursor, (size_t)ret);
512 }
513
514 dout("%s %p msg %p done\n", __func__, con, msg);
515
516 /* prepare and queue up footer, too */
517 if (do_datacrc)
518 msg->footer.data_crc = cpu_to_le32(crc);
519 else
520 msg->footer.flags |= CEPH_MSG_FOOTER_NOCRC;
521 con_out_kvec_reset(con);
522 prepare_write_message_footer(con);
523
524 return 1; /* must return > 0 to indicate success */
525 }
526
527 /*
528 * write some zeros
529 */
write_partial_skip(struct ceph_connection * con)530 static int write_partial_skip(struct ceph_connection *con)
531 {
532 int more = MSG_MORE | MSG_SENDPAGE_NOTLAST;
533 int ret;
534
535 dout("%s %p %d left\n", __func__, con, con->v1.out_skip);
536 while (con->v1.out_skip > 0) {
537 size_t size = min(con->v1.out_skip, (int)PAGE_SIZE);
538
539 if (size == con->v1.out_skip)
540 more = MSG_MORE;
541 ret = ceph_tcp_sendpage(con->sock, ceph_zero_page, 0, size,
542 more);
543 if (ret <= 0)
544 goto out;
545 con->v1.out_skip -= ret;
546 }
547 ret = 1;
548 out:
549 return ret;
550 }
551
552 /*
553 * Prepare to read connection handshake, or an ack.
554 */
prepare_read_banner(struct ceph_connection * con)555 static void prepare_read_banner(struct ceph_connection *con)
556 {
557 dout("prepare_read_banner %p\n", con);
558 con->v1.in_base_pos = 0;
559 }
560
prepare_read_connect(struct ceph_connection * con)561 static void prepare_read_connect(struct ceph_connection *con)
562 {
563 dout("prepare_read_connect %p\n", con);
564 con->v1.in_base_pos = 0;
565 }
566
prepare_read_ack(struct ceph_connection * con)567 static void prepare_read_ack(struct ceph_connection *con)
568 {
569 dout("prepare_read_ack %p\n", con);
570 con->v1.in_base_pos = 0;
571 }
572
prepare_read_seq(struct ceph_connection * con)573 static void prepare_read_seq(struct ceph_connection *con)
574 {
575 dout("prepare_read_seq %p\n", con);
576 con->v1.in_base_pos = 0;
577 con->v1.in_tag = CEPH_MSGR_TAG_SEQ;
578 }
579
prepare_read_tag(struct ceph_connection * con)580 static void prepare_read_tag(struct ceph_connection *con)
581 {
582 dout("prepare_read_tag %p\n", con);
583 con->v1.in_base_pos = 0;
584 con->v1.in_tag = CEPH_MSGR_TAG_READY;
585 }
586
prepare_read_keepalive_ack(struct ceph_connection * con)587 static void prepare_read_keepalive_ack(struct ceph_connection *con)
588 {
589 dout("prepare_read_keepalive_ack %p\n", con);
590 con->v1.in_base_pos = 0;
591 }
592
593 /*
594 * Prepare to read a message.
595 */
prepare_read_message(struct ceph_connection * con)596 static int prepare_read_message(struct ceph_connection *con)
597 {
598 dout("prepare_read_message %p\n", con);
599 BUG_ON(con->in_msg != NULL);
600 con->v1.in_base_pos = 0;
601 con->in_front_crc = con->in_middle_crc = con->in_data_crc = 0;
602 return 0;
603 }
604
read_partial(struct ceph_connection * con,int end,int size,void * object)605 static int read_partial(struct ceph_connection *con,
606 int end, int size, void *object)
607 {
608 while (con->v1.in_base_pos < end) {
609 int left = end - con->v1.in_base_pos;
610 int have = size - left;
611 int ret = ceph_tcp_recvmsg(con->sock, object + have, left);
612 if (ret <= 0)
613 return ret;
614 con->v1.in_base_pos += ret;
615 }
616 return 1;
617 }
618
619 /*
620 * Read all or part of the connect-side handshake on a new connection
621 */
read_partial_banner(struct ceph_connection * con)622 static int read_partial_banner(struct ceph_connection *con)
623 {
624 int size;
625 int end;
626 int ret;
627
628 dout("read_partial_banner %p at %d\n", con, con->v1.in_base_pos);
629
630 /* peer's banner */
631 size = strlen(CEPH_BANNER);
632 end = size;
633 ret = read_partial(con, end, size, con->v1.in_banner);
634 if (ret <= 0)
635 goto out;
636
637 size = sizeof(con->v1.actual_peer_addr);
638 end += size;
639 ret = read_partial(con, end, size, &con->v1.actual_peer_addr);
640 if (ret <= 0)
641 goto out;
642 ceph_decode_banner_addr(&con->v1.actual_peer_addr);
643
644 size = sizeof(con->v1.peer_addr_for_me);
645 end += size;
646 ret = read_partial(con, end, size, &con->v1.peer_addr_for_me);
647 if (ret <= 0)
648 goto out;
649 ceph_decode_banner_addr(&con->v1.peer_addr_for_me);
650
651 out:
652 return ret;
653 }
654
read_partial_connect(struct ceph_connection * con)655 static int read_partial_connect(struct ceph_connection *con)
656 {
657 int size;
658 int end;
659 int ret;
660
661 dout("read_partial_connect %p at %d\n", con, con->v1.in_base_pos);
662
663 size = sizeof(con->v1.in_reply);
664 end = size;
665 ret = read_partial(con, end, size, &con->v1.in_reply);
666 if (ret <= 0)
667 goto out;
668
669 if (con->v1.auth) {
670 size = le32_to_cpu(con->v1.in_reply.authorizer_len);
671 if (size > con->v1.auth->authorizer_reply_buf_len) {
672 pr_err("authorizer reply too big: %d > %zu\n", size,
673 con->v1.auth->authorizer_reply_buf_len);
674 ret = -EINVAL;
675 goto out;
676 }
677
678 end += size;
679 ret = read_partial(con, end, size,
680 con->v1.auth->authorizer_reply_buf);
681 if (ret <= 0)
682 goto out;
683 }
684
685 dout("read_partial_connect %p tag %d, con_seq = %u, g_seq = %u\n",
686 con, con->v1.in_reply.tag,
687 le32_to_cpu(con->v1.in_reply.connect_seq),
688 le32_to_cpu(con->v1.in_reply.global_seq));
689 out:
690 return ret;
691 }
692
693 /*
694 * Verify the hello banner looks okay.
695 */
verify_hello(struct ceph_connection * con)696 static int verify_hello(struct ceph_connection *con)
697 {
698 if (memcmp(con->v1.in_banner, CEPH_BANNER, strlen(CEPH_BANNER))) {
699 pr_err("connect to %s got bad banner\n",
700 ceph_pr_addr(&con->peer_addr));
701 con->error_msg = "protocol error, bad banner";
702 return -1;
703 }
704 return 0;
705 }
706
process_banner(struct ceph_connection * con)707 static int process_banner(struct ceph_connection *con)
708 {
709 struct ceph_entity_addr *my_addr = &con->msgr->inst.addr;
710
711 dout("process_banner on %p\n", con);
712
713 if (verify_hello(con) < 0)
714 return -1;
715
716 /*
717 * Make sure the other end is who we wanted. note that the other
718 * end may not yet know their ip address, so if it's 0.0.0.0, give
719 * them the benefit of the doubt.
720 */
721 if (memcmp(&con->peer_addr, &con->v1.actual_peer_addr,
722 sizeof(con->peer_addr)) != 0 &&
723 !(ceph_addr_is_blank(&con->v1.actual_peer_addr) &&
724 con->v1.actual_peer_addr.nonce == con->peer_addr.nonce)) {
725 pr_warn("wrong peer, want %s/%u, got %s/%u\n",
726 ceph_pr_addr(&con->peer_addr),
727 le32_to_cpu(con->peer_addr.nonce),
728 ceph_pr_addr(&con->v1.actual_peer_addr),
729 le32_to_cpu(con->v1.actual_peer_addr.nonce));
730 con->error_msg = "wrong peer at address";
731 return -1;
732 }
733
734 /*
735 * did we learn our address?
736 */
737 if (ceph_addr_is_blank(my_addr)) {
738 memcpy(&my_addr->in_addr,
739 &con->v1.peer_addr_for_me.in_addr,
740 sizeof(con->v1.peer_addr_for_me.in_addr));
741 ceph_addr_set_port(my_addr, 0);
742 ceph_encode_my_addr(con->msgr);
743 dout("process_banner learned my addr is %s\n",
744 ceph_pr_addr(my_addr));
745 }
746
747 return 0;
748 }
749
process_connect(struct ceph_connection * con)750 static int process_connect(struct ceph_connection *con)
751 {
752 u64 sup_feat = from_msgr(con->msgr)->supported_features;
753 u64 req_feat = from_msgr(con->msgr)->required_features;
754 u64 server_feat = le64_to_cpu(con->v1.in_reply.features);
755 int ret;
756
757 dout("process_connect on %p tag %d\n", con, con->v1.in_tag);
758
759 if (con->v1.auth) {
760 int len = le32_to_cpu(con->v1.in_reply.authorizer_len);
761
762 /*
763 * Any connection that defines ->get_authorizer()
764 * should also define ->add_authorizer_challenge() and
765 * ->verify_authorizer_reply().
766 *
767 * See get_connect_authorizer().
768 */
769 if (con->v1.in_reply.tag ==
770 CEPH_MSGR_TAG_CHALLENGE_AUTHORIZER) {
771 ret = con->ops->add_authorizer_challenge(
772 con, con->v1.auth->authorizer_reply_buf, len);
773 if (ret < 0)
774 return ret;
775
776 con_out_kvec_reset(con);
777 __prepare_write_connect(con);
778 prepare_read_connect(con);
779 return 0;
780 }
781
782 if (len) {
783 ret = con->ops->verify_authorizer_reply(con);
784 if (ret < 0) {
785 con->error_msg = "bad authorize reply";
786 return ret;
787 }
788 }
789 }
790
791 switch (con->v1.in_reply.tag) {
792 case CEPH_MSGR_TAG_FEATURES:
793 pr_err("%s%lld %s feature set mismatch,"
794 " my %llx < server's %llx, missing %llx\n",
795 ENTITY_NAME(con->peer_name),
796 ceph_pr_addr(&con->peer_addr),
797 sup_feat, server_feat, server_feat & ~sup_feat);
798 con->error_msg = "missing required protocol features";
799 return -1;
800
801 case CEPH_MSGR_TAG_BADPROTOVER:
802 pr_err("%s%lld %s protocol version mismatch,"
803 " my %d != server's %d\n",
804 ENTITY_NAME(con->peer_name),
805 ceph_pr_addr(&con->peer_addr),
806 le32_to_cpu(con->v1.out_connect.protocol_version),
807 le32_to_cpu(con->v1.in_reply.protocol_version));
808 con->error_msg = "protocol version mismatch";
809 return -1;
810
811 case CEPH_MSGR_TAG_BADAUTHORIZER:
812 con->v1.auth_retry++;
813 dout("process_connect %p got BADAUTHORIZER attempt %d\n", con,
814 con->v1.auth_retry);
815 if (con->v1.auth_retry == 2) {
816 con->error_msg = "connect authorization failure";
817 return -1;
818 }
819 con_out_kvec_reset(con);
820 ret = prepare_write_connect(con);
821 if (ret < 0)
822 return ret;
823 prepare_read_connect(con);
824 break;
825
826 case CEPH_MSGR_TAG_RESETSESSION:
827 /*
828 * If we connected with a large connect_seq but the peer
829 * has no record of a session with us (no connection, or
830 * connect_seq == 0), they will send RESETSESION to indicate
831 * that they must have reset their session, and may have
832 * dropped messages.
833 */
834 dout("process_connect got RESET peer seq %u\n",
835 le32_to_cpu(con->v1.in_reply.connect_seq));
836 pr_info("%s%lld %s session reset\n",
837 ENTITY_NAME(con->peer_name),
838 ceph_pr_addr(&con->peer_addr));
839 ceph_con_reset_session(con);
840 con_out_kvec_reset(con);
841 ret = prepare_write_connect(con);
842 if (ret < 0)
843 return ret;
844 prepare_read_connect(con);
845
846 /* Tell ceph about it. */
847 mutex_unlock(&con->mutex);
848 if (con->ops->peer_reset)
849 con->ops->peer_reset(con);
850 mutex_lock(&con->mutex);
851 if (con->state != CEPH_CON_S_V1_CONNECT_MSG)
852 return -EAGAIN;
853 break;
854
855 case CEPH_MSGR_TAG_RETRY_SESSION:
856 /*
857 * If we sent a smaller connect_seq than the peer has, try
858 * again with a larger value.
859 */
860 dout("process_connect got RETRY_SESSION my seq %u, peer %u\n",
861 le32_to_cpu(con->v1.out_connect.connect_seq),
862 le32_to_cpu(con->v1.in_reply.connect_seq));
863 con->v1.connect_seq = le32_to_cpu(con->v1.in_reply.connect_seq);
864 con_out_kvec_reset(con);
865 ret = prepare_write_connect(con);
866 if (ret < 0)
867 return ret;
868 prepare_read_connect(con);
869 break;
870
871 case CEPH_MSGR_TAG_RETRY_GLOBAL:
872 /*
873 * If we sent a smaller global_seq than the peer has, try
874 * again with a larger value.
875 */
876 dout("process_connect got RETRY_GLOBAL my %u peer_gseq %u\n",
877 con->v1.peer_global_seq,
878 le32_to_cpu(con->v1.in_reply.global_seq));
879 ceph_get_global_seq(con->msgr,
880 le32_to_cpu(con->v1.in_reply.global_seq));
881 con_out_kvec_reset(con);
882 ret = prepare_write_connect(con);
883 if (ret < 0)
884 return ret;
885 prepare_read_connect(con);
886 break;
887
888 case CEPH_MSGR_TAG_SEQ:
889 case CEPH_MSGR_TAG_READY:
890 if (req_feat & ~server_feat) {
891 pr_err("%s%lld %s protocol feature mismatch,"
892 " my required %llx > server's %llx, need %llx\n",
893 ENTITY_NAME(con->peer_name),
894 ceph_pr_addr(&con->peer_addr),
895 req_feat, server_feat, req_feat & ~server_feat);
896 con->error_msg = "missing required protocol features";
897 return -1;
898 }
899
900 WARN_ON(con->state != CEPH_CON_S_V1_CONNECT_MSG);
901 con->state = CEPH_CON_S_OPEN;
902 con->v1.auth_retry = 0; /* we authenticated; clear flag */
903 con->v1.peer_global_seq =
904 le32_to_cpu(con->v1.in_reply.global_seq);
905 con->v1.connect_seq++;
906 con->peer_features = server_feat;
907 dout("process_connect got READY gseq %d cseq %d (%d)\n",
908 con->v1.peer_global_seq,
909 le32_to_cpu(con->v1.in_reply.connect_seq),
910 con->v1.connect_seq);
911 WARN_ON(con->v1.connect_seq !=
912 le32_to_cpu(con->v1.in_reply.connect_seq));
913
914 if (con->v1.in_reply.flags & CEPH_MSG_CONNECT_LOSSY)
915 ceph_con_flag_set(con, CEPH_CON_F_LOSSYTX);
916
917 con->delay = 0; /* reset backoff memory */
918
919 if (con->v1.in_reply.tag == CEPH_MSGR_TAG_SEQ) {
920 prepare_write_seq(con);
921 prepare_read_seq(con);
922 } else {
923 prepare_read_tag(con);
924 }
925 break;
926
927 case CEPH_MSGR_TAG_WAIT:
928 /*
929 * If there is a connection race (we are opening
930 * connections to each other), one of us may just have
931 * to WAIT. This shouldn't happen if we are the
932 * client.
933 */
934 con->error_msg = "protocol error, got WAIT as client";
935 return -1;
936
937 default:
938 con->error_msg = "protocol error, garbage tag during connect";
939 return -1;
940 }
941 return 0;
942 }
943
944 /*
945 * read (part of) an ack
946 */
read_partial_ack(struct ceph_connection * con)947 static int read_partial_ack(struct ceph_connection *con)
948 {
949 int size = sizeof(con->v1.in_temp_ack);
950 int end = size;
951
952 return read_partial(con, end, size, &con->v1.in_temp_ack);
953 }
954
955 /*
956 * We can finally discard anything that's been acked.
957 */
process_ack(struct ceph_connection * con)958 static void process_ack(struct ceph_connection *con)
959 {
960 u64 ack = le64_to_cpu(con->v1.in_temp_ack);
961
962 if (con->v1.in_tag == CEPH_MSGR_TAG_ACK)
963 ceph_con_discard_sent(con, ack);
964 else
965 ceph_con_discard_requeued(con, ack);
966
967 prepare_read_tag(con);
968 }
969
read_partial_message_section(struct ceph_connection * con,struct kvec * section,unsigned int sec_len,u32 * crc)970 static int read_partial_message_section(struct ceph_connection *con,
971 struct kvec *section,
972 unsigned int sec_len, u32 *crc)
973 {
974 int ret, left;
975
976 BUG_ON(!section);
977
978 while (section->iov_len < sec_len) {
979 BUG_ON(section->iov_base == NULL);
980 left = sec_len - section->iov_len;
981 ret = ceph_tcp_recvmsg(con->sock, (char *)section->iov_base +
982 section->iov_len, left);
983 if (ret <= 0)
984 return ret;
985 section->iov_len += ret;
986 }
987 if (section->iov_len == sec_len)
988 *crc = crc32c(0, section->iov_base, section->iov_len);
989
990 return 1;
991 }
992
read_partial_msg_data(struct ceph_connection * con)993 static int read_partial_msg_data(struct ceph_connection *con)
994 {
995 struct ceph_msg_data_cursor *cursor = &con->in_msg->cursor;
996 bool do_datacrc = !ceph_test_opt(from_msgr(con->msgr), NOCRC);
997 struct page *page;
998 size_t page_offset;
999 size_t length;
1000 u32 crc = 0;
1001 int ret;
1002
1003 if (do_datacrc)
1004 crc = con->in_data_crc;
1005 while (cursor->total_resid) {
1006 if (!cursor->resid) {
1007 ceph_msg_data_advance(cursor, 0);
1008 continue;
1009 }
1010
1011 page = ceph_msg_data_next(cursor, &page_offset, &length, NULL);
1012 ret = ceph_tcp_recvpage(con->sock, page, page_offset, length);
1013 if (ret <= 0) {
1014 if (do_datacrc)
1015 con->in_data_crc = crc;
1016
1017 return ret;
1018 }
1019
1020 if (do_datacrc)
1021 crc = ceph_crc32c_page(crc, page, page_offset, ret);
1022 ceph_msg_data_advance(cursor, (size_t)ret);
1023 }
1024 if (do_datacrc)
1025 con->in_data_crc = crc;
1026
1027 return 1; /* must return > 0 to indicate success */
1028 }
1029
read_partial_msg_data_bounce(struct ceph_connection * con)1030 static int read_partial_msg_data_bounce(struct ceph_connection *con)
1031 {
1032 struct ceph_msg_data_cursor *cursor = &con->in_msg->cursor;
1033 struct page *page;
1034 size_t off, len;
1035 u32 crc;
1036 int ret;
1037
1038 if (unlikely(!con->bounce_page)) {
1039 con->bounce_page = alloc_page(GFP_NOIO);
1040 if (!con->bounce_page) {
1041 pr_err("failed to allocate bounce page\n");
1042 return -ENOMEM;
1043 }
1044 }
1045
1046 crc = con->in_data_crc;
1047 while (cursor->total_resid) {
1048 if (!cursor->resid) {
1049 ceph_msg_data_advance(cursor, 0);
1050 continue;
1051 }
1052
1053 page = ceph_msg_data_next(cursor, &off, &len, NULL);
1054 ret = ceph_tcp_recvpage(con->sock, con->bounce_page, 0, len);
1055 if (ret <= 0) {
1056 con->in_data_crc = crc;
1057 return ret;
1058 }
1059
1060 crc = crc32c(crc, page_address(con->bounce_page), ret);
1061 memcpy_to_page(page, off, page_address(con->bounce_page), ret);
1062
1063 ceph_msg_data_advance(cursor, ret);
1064 }
1065 con->in_data_crc = crc;
1066
1067 return 1; /* must return > 0 to indicate success */
1068 }
1069
1070 /*
1071 * read (part of) a message.
1072 */
read_partial_message(struct ceph_connection * con)1073 static int read_partial_message(struct ceph_connection *con)
1074 {
1075 struct ceph_msg *m = con->in_msg;
1076 int size;
1077 int end;
1078 int ret;
1079 unsigned int front_len, middle_len, data_len;
1080 bool do_datacrc = !ceph_test_opt(from_msgr(con->msgr), NOCRC);
1081 bool need_sign = (con->peer_features & CEPH_FEATURE_MSG_AUTH);
1082 u64 seq;
1083 u32 crc;
1084
1085 dout("read_partial_message con %p msg %p\n", con, m);
1086
1087 /* header */
1088 size = sizeof(con->v1.in_hdr);
1089 end = size;
1090 ret = read_partial(con, end, size, &con->v1.in_hdr);
1091 if (ret <= 0)
1092 return ret;
1093
1094 crc = crc32c(0, &con->v1.in_hdr, offsetof(struct ceph_msg_header, crc));
1095 if (cpu_to_le32(crc) != con->v1.in_hdr.crc) {
1096 pr_err("read_partial_message bad hdr crc %u != expected %u\n",
1097 crc, con->v1.in_hdr.crc);
1098 return -EBADMSG;
1099 }
1100
1101 front_len = le32_to_cpu(con->v1.in_hdr.front_len);
1102 if (front_len > CEPH_MSG_MAX_FRONT_LEN)
1103 return -EIO;
1104 middle_len = le32_to_cpu(con->v1.in_hdr.middle_len);
1105 if (middle_len > CEPH_MSG_MAX_MIDDLE_LEN)
1106 return -EIO;
1107 data_len = le32_to_cpu(con->v1.in_hdr.data_len);
1108 if (data_len > CEPH_MSG_MAX_DATA_LEN)
1109 return -EIO;
1110
1111 /* verify seq# */
1112 seq = le64_to_cpu(con->v1.in_hdr.seq);
1113 if ((s64)seq - (s64)con->in_seq < 1) {
1114 pr_info("skipping %s%lld %s seq %lld expected %lld\n",
1115 ENTITY_NAME(con->peer_name),
1116 ceph_pr_addr(&con->peer_addr),
1117 seq, con->in_seq + 1);
1118 con->v1.in_base_pos = -front_len - middle_len - data_len -
1119 sizeof_footer(con);
1120 con->v1.in_tag = CEPH_MSGR_TAG_READY;
1121 return 1;
1122 } else if ((s64)seq - (s64)con->in_seq > 1) {
1123 pr_err("read_partial_message bad seq %lld expected %lld\n",
1124 seq, con->in_seq + 1);
1125 con->error_msg = "bad message sequence # for incoming message";
1126 return -EBADE;
1127 }
1128
1129 /* allocate message? */
1130 if (!con->in_msg) {
1131 int skip = 0;
1132
1133 dout("got hdr type %d front %d data %d\n", con->v1.in_hdr.type,
1134 front_len, data_len);
1135 ret = ceph_con_in_msg_alloc(con, &con->v1.in_hdr, &skip);
1136 if (ret < 0)
1137 return ret;
1138
1139 BUG_ON((!con->in_msg) ^ skip);
1140 if (skip) {
1141 /* skip this message */
1142 dout("alloc_msg said skip message\n");
1143 con->v1.in_base_pos = -front_len - middle_len -
1144 data_len - sizeof_footer(con);
1145 con->v1.in_tag = CEPH_MSGR_TAG_READY;
1146 con->in_seq++;
1147 return 1;
1148 }
1149
1150 BUG_ON(!con->in_msg);
1151 BUG_ON(con->in_msg->con != con);
1152 m = con->in_msg;
1153 m->front.iov_len = 0; /* haven't read it yet */
1154 if (m->middle)
1155 m->middle->vec.iov_len = 0;
1156
1157 /* prepare for data payload, if any */
1158
1159 if (data_len)
1160 prepare_message_data(con->in_msg, data_len);
1161 }
1162
1163 /* front */
1164 ret = read_partial_message_section(con, &m->front, front_len,
1165 &con->in_front_crc);
1166 if (ret <= 0)
1167 return ret;
1168
1169 /* middle */
1170 if (m->middle) {
1171 ret = read_partial_message_section(con, &m->middle->vec,
1172 middle_len,
1173 &con->in_middle_crc);
1174 if (ret <= 0)
1175 return ret;
1176 }
1177
1178 /* (page) data */
1179 if (data_len) {
1180 if (!m->num_data_items)
1181 return -EIO;
1182
1183 if (ceph_test_opt(from_msgr(con->msgr), RXBOUNCE))
1184 ret = read_partial_msg_data_bounce(con);
1185 else
1186 ret = read_partial_msg_data(con);
1187 if (ret <= 0)
1188 return ret;
1189 }
1190
1191 /* footer */
1192 size = sizeof_footer(con);
1193 end += size;
1194 ret = read_partial(con, end, size, &m->footer);
1195 if (ret <= 0)
1196 return ret;
1197
1198 if (!need_sign) {
1199 m->footer.flags = m->old_footer.flags;
1200 m->footer.sig = 0;
1201 }
1202
1203 dout("read_partial_message got msg %p %d (%u) + %d (%u) + %d (%u)\n",
1204 m, front_len, m->footer.front_crc, middle_len,
1205 m->footer.middle_crc, data_len, m->footer.data_crc);
1206
1207 /* crc ok? */
1208 if (con->in_front_crc != le32_to_cpu(m->footer.front_crc)) {
1209 pr_err("read_partial_message %p front crc %u != exp. %u\n",
1210 m, con->in_front_crc, m->footer.front_crc);
1211 return -EBADMSG;
1212 }
1213 if (con->in_middle_crc != le32_to_cpu(m->footer.middle_crc)) {
1214 pr_err("read_partial_message %p middle crc %u != exp %u\n",
1215 m, con->in_middle_crc, m->footer.middle_crc);
1216 return -EBADMSG;
1217 }
1218 if (do_datacrc &&
1219 (m->footer.flags & CEPH_MSG_FOOTER_NOCRC) == 0 &&
1220 con->in_data_crc != le32_to_cpu(m->footer.data_crc)) {
1221 pr_err("read_partial_message %p data crc %u != exp. %u\n", m,
1222 con->in_data_crc, le32_to_cpu(m->footer.data_crc));
1223 return -EBADMSG;
1224 }
1225
1226 if (need_sign && con->ops->check_message_signature &&
1227 con->ops->check_message_signature(m)) {
1228 pr_err("read_partial_message %p signature check failed\n", m);
1229 return -EBADMSG;
1230 }
1231
1232 return 1; /* done! */
1233 }
1234
read_keepalive_ack(struct ceph_connection * con)1235 static int read_keepalive_ack(struct ceph_connection *con)
1236 {
1237 struct ceph_timespec ceph_ts;
1238 size_t size = sizeof(ceph_ts);
1239 int ret = read_partial(con, size, size, &ceph_ts);
1240 if (ret <= 0)
1241 return ret;
1242 ceph_decode_timespec64(&con->last_keepalive_ack, &ceph_ts);
1243 prepare_read_tag(con);
1244 return 1;
1245 }
1246
1247 /*
1248 * Read what we can from the socket.
1249 */
ceph_con_v1_try_read(struct ceph_connection * con)1250 int ceph_con_v1_try_read(struct ceph_connection *con)
1251 {
1252 int ret = -1;
1253
1254 more:
1255 dout("try_read start %p state %d\n", con, con->state);
1256 if (con->state != CEPH_CON_S_V1_BANNER &&
1257 con->state != CEPH_CON_S_V1_CONNECT_MSG &&
1258 con->state != CEPH_CON_S_OPEN)
1259 return 0;
1260
1261 BUG_ON(!con->sock);
1262
1263 dout("try_read tag %d in_base_pos %d\n", con->v1.in_tag,
1264 con->v1.in_base_pos);
1265
1266 if (con->state == CEPH_CON_S_V1_BANNER) {
1267 ret = read_partial_banner(con);
1268 if (ret <= 0)
1269 goto out;
1270 ret = process_banner(con);
1271 if (ret < 0)
1272 goto out;
1273
1274 con->state = CEPH_CON_S_V1_CONNECT_MSG;
1275
1276 /*
1277 * Received banner is good, exchange connection info.
1278 * Do not reset out_kvec, as sending our banner raced
1279 * with receiving peer banner after connect completed.
1280 */
1281 ret = prepare_write_connect(con);
1282 if (ret < 0)
1283 goto out;
1284 prepare_read_connect(con);
1285
1286 /* Send connection info before awaiting response */
1287 goto out;
1288 }
1289
1290 if (con->state == CEPH_CON_S_V1_CONNECT_MSG) {
1291 ret = read_partial_connect(con);
1292 if (ret <= 0)
1293 goto out;
1294 ret = process_connect(con);
1295 if (ret < 0)
1296 goto out;
1297 goto more;
1298 }
1299
1300 WARN_ON(con->state != CEPH_CON_S_OPEN);
1301
1302 if (con->v1.in_base_pos < 0) {
1303 /*
1304 * skipping + discarding content.
1305 */
1306 ret = ceph_tcp_recvmsg(con->sock, NULL, -con->v1.in_base_pos);
1307 if (ret <= 0)
1308 goto out;
1309 dout("skipped %d / %d bytes\n", ret, -con->v1.in_base_pos);
1310 con->v1.in_base_pos += ret;
1311 if (con->v1.in_base_pos)
1312 goto more;
1313 }
1314 if (con->v1.in_tag == CEPH_MSGR_TAG_READY) {
1315 /*
1316 * what's next?
1317 */
1318 ret = ceph_tcp_recvmsg(con->sock, &con->v1.in_tag, 1);
1319 if (ret <= 0)
1320 goto out;
1321 dout("try_read got tag %d\n", con->v1.in_tag);
1322 switch (con->v1.in_tag) {
1323 case CEPH_MSGR_TAG_MSG:
1324 prepare_read_message(con);
1325 break;
1326 case CEPH_MSGR_TAG_ACK:
1327 prepare_read_ack(con);
1328 break;
1329 case CEPH_MSGR_TAG_KEEPALIVE2_ACK:
1330 prepare_read_keepalive_ack(con);
1331 break;
1332 case CEPH_MSGR_TAG_CLOSE:
1333 ceph_con_close_socket(con);
1334 con->state = CEPH_CON_S_CLOSED;
1335 goto out;
1336 default:
1337 goto bad_tag;
1338 }
1339 }
1340 if (con->v1.in_tag == CEPH_MSGR_TAG_MSG) {
1341 ret = read_partial_message(con);
1342 if (ret <= 0) {
1343 switch (ret) {
1344 case -EBADMSG:
1345 con->error_msg = "bad crc/signature";
1346 fallthrough;
1347 case -EBADE:
1348 ret = -EIO;
1349 break;
1350 case -EIO:
1351 con->error_msg = "io error";
1352 break;
1353 }
1354 goto out;
1355 }
1356 if (con->v1.in_tag == CEPH_MSGR_TAG_READY)
1357 goto more;
1358 ceph_con_process_message(con);
1359 if (con->state == CEPH_CON_S_OPEN)
1360 prepare_read_tag(con);
1361 goto more;
1362 }
1363 if (con->v1.in_tag == CEPH_MSGR_TAG_ACK ||
1364 con->v1.in_tag == CEPH_MSGR_TAG_SEQ) {
1365 /*
1366 * the final handshake seq exchange is semantically
1367 * equivalent to an ACK
1368 */
1369 ret = read_partial_ack(con);
1370 if (ret <= 0)
1371 goto out;
1372 process_ack(con);
1373 goto more;
1374 }
1375 if (con->v1.in_tag == CEPH_MSGR_TAG_KEEPALIVE2_ACK) {
1376 ret = read_keepalive_ack(con);
1377 if (ret <= 0)
1378 goto out;
1379 goto more;
1380 }
1381
1382 out:
1383 dout("try_read done on %p ret %d\n", con, ret);
1384 return ret;
1385
1386 bad_tag:
1387 pr_err("try_read bad tag %d\n", con->v1.in_tag);
1388 con->error_msg = "protocol error, garbage tag";
1389 ret = -1;
1390 goto out;
1391 }
1392
1393 /*
1394 * Write something to the socket. Called in a worker thread when the
1395 * socket appears to be writeable and we have something ready to send.
1396 */
ceph_con_v1_try_write(struct ceph_connection * con)1397 int ceph_con_v1_try_write(struct ceph_connection *con)
1398 {
1399 int ret = 1;
1400
1401 dout("try_write start %p state %d\n", con, con->state);
1402 if (con->state != CEPH_CON_S_PREOPEN &&
1403 con->state != CEPH_CON_S_V1_BANNER &&
1404 con->state != CEPH_CON_S_V1_CONNECT_MSG &&
1405 con->state != CEPH_CON_S_OPEN)
1406 return 0;
1407
1408 /* open the socket first? */
1409 if (con->state == CEPH_CON_S_PREOPEN) {
1410 BUG_ON(con->sock);
1411 con->state = CEPH_CON_S_V1_BANNER;
1412
1413 con_out_kvec_reset(con);
1414 prepare_write_banner(con);
1415 prepare_read_banner(con);
1416
1417 BUG_ON(con->in_msg);
1418 con->v1.in_tag = CEPH_MSGR_TAG_READY;
1419 dout("try_write initiating connect on %p new state %d\n",
1420 con, con->state);
1421 ret = ceph_tcp_connect(con);
1422 if (ret < 0) {
1423 con->error_msg = "connect error";
1424 goto out;
1425 }
1426 }
1427
1428 more:
1429 dout("try_write out_kvec_bytes %d\n", con->v1.out_kvec_bytes);
1430 BUG_ON(!con->sock);
1431
1432 /* kvec data queued? */
1433 if (con->v1.out_kvec_left) {
1434 ret = write_partial_kvec(con);
1435 if (ret <= 0)
1436 goto out;
1437 }
1438 if (con->v1.out_skip) {
1439 ret = write_partial_skip(con);
1440 if (ret <= 0)
1441 goto out;
1442 }
1443
1444 /* msg pages? */
1445 if (con->out_msg) {
1446 if (con->v1.out_msg_done) {
1447 ceph_msg_put(con->out_msg);
1448 con->out_msg = NULL; /* we're done with this one */
1449 goto do_next;
1450 }
1451
1452 ret = write_partial_message_data(con);
1453 if (ret == 1)
1454 goto more; /* we need to send the footer, too! */
1455 if (ret == 0)
1456 goto out;
1457 if (ret < 0) {
1458 dout("try_write write_partial_message_data err %d\n",
1459 ret);
1460 goto out;
1461 }
1462 }
1463
1464 do_next:
1465 if (con->state == CEPH_CON_S_OPEN) {
1466 if (ceph_con_flag_test_and_clear(con,
1467 CEPH_CON_F_KEEPALIVE_PENDING)) {
1468 prepare_write_keepalive(con);
1469 goto more;
1470 }
1471 /* is anything else pending? */
1472 if (!list_empty(&con->out_queue)) {
1473 prepare_write_message(con);
1474 goto more;
1475 }
1476 if (con->in_seq > con->in_seq_acked) {
1477 prepare_write_ack(con);
1478 goto more;
1479 }
1480 }
1481
1482 /* Nothing to do! */
1483 ceph_con_flag_clear(con, CEPH_CON_F_WRITE_PENDING);
1484 dout("try_write nothing else to write.\n");
1485 ret = 0;
1486 out:
1487 dout("try_write done on %p ret %d\n", con, ret);
1488 return ret;
1489 }
1490
ceph_con_v1_revoke(struct ceph_connection * con)1491 void ceph_con_v1_revoke(struct ceph_connection *con)
1492 {
1493 struct ceph_msg *msg = con->out_msg;
1494
1495 WARN_ON(con->v1.out_skip);
1496 /* footer */
1497 if (con->v1.out_msg_done) {
1498 con->v1.out_skip += con_out_kvec_skip(con);
1499 } else {
1500 WARN_ON(!msg->data_length);
1501 con->v1.out_skip += sizeof_footer(con);
1502 }
1503 /* data, middle, front */
1504 if (msg->data_length)
1505 con->v1.out_skip += msg->cursor.total_resid;
1506 if (msg->middle)
1507 con->v1.out_skip += con_out_kvec_skip(con);
1508 con->v1.out_skip += con_out_kvec_skip(con);
1509
1510 dout("%s con %p out_kvec_bytes %d out_skip %d\n", __func__, con,
1511 con->v1.out_kvec_bytes, con->v1.out_skip);
1512 }
1513
ceph_con_v1_revoke_incoming(struct ceph_connection * con)1514 void ceph_con_v1_revoke_incoming(struct ceph_connection *con)
1515 {
1516 unsigned int front_len = le32_to_cpu(con->v1.in_hdr.front_len);
1517 unsigned int middle_len = le32_to_cpu(con->v1.in_hdr.middle_len);
1518 unsigned int data_len = le32_to_cpu(con->v1.in_hdr.data_len);
1519
1520 /* skip rest of message */
1521 con->v1.in_base_pos = con->v1.in_base_pos -
1522 sizeof(struct ceph_msg_header) -
1523 front_len -
1524 middle_len -
1525 data_len -
1526 sizeof(struct ceph_msg_footer);
1527
1528 con->v1.in_tag = CEPH_MSGR_TAG_READY;
1529 con->in_seq++;
1530
1531 dout("%s con %p in_base_pos %d\n", __func__, con, con->v1.in_base_pos);
1532 }
1533
ceph_con_v1_opened(struct ceph_connection * con)1534 bool ceph_con_v1_opened(struct ceph_connection *con)
1535 {
1536 return con->v1.connect_seq;
1537 }
1538
ceph_con_v1_reset_session(struct ceph_connection * con)1539 void ceph_con_v1_reset_session(struct ceph_connection *con)
1540 {
1541 con->v1.connect_seq = 0;
1542 con->v1.peer_global_seq = 0;
1543 }
1544
ceph_con_v1_reset_protocol(struct ceph_connection * con)1545 void ceph_con_v1_reset_protocol(struct ceph_connection *con)
1546 {
1547 con->v1.out_skip = 0;
1548 }
1549