aboutsummaryrefslogtreecommitdiffstats
path: root/net/ceph/osd_client.c
diff options
context:
space:
mode:
authorAlex Elder <elder@inktank.com>2012-11-13 22:11:15 -0500
committerAlex Elder <elder@inktank.com>2013-01-17 16:52:03 -0500
commitc885837f7d4f8c4f5cb2a744cc6929bc078e9dc0 (patch)
tree2b4e60e8e7aee98e8f4e8c56119978160ba85cfd /net/ceph/osd_client.c
parent7c3d22cf16f1bbcb37a73e88338c042bb49ff112 (diff)
libceph: always allow trail in osd request
An osd request structure contains an optional trail portion, which if present will contain data to be passed in the payload portion of the message containing the request. The trail field is a ceph_pagelist pointer, and if null it indicates there is no trail. A ceph_pagelist structure contains a length field, and it can legitimately hold value 0. Make use of this to change the interpretation of the "trail" of an osd request so that every osd request has trailing data, it just might have length 0. This means we change the r_trail field in a ceph_osd_request structure from a pointer to a structure that is always initialized. Note that in ceph_osdc_start_request(), the trail pointer (or now address of that structure) is assigned to a ceph message's trail field. Here's why that's still OK (looking at net/ceph/messenger.c): - What would have resulted in a null pointer previously will now refer to a 0-length page list. That message trail pointer is used in two functions, write_partial_msg_pages() and out_msg_pos_next(). - In write_partial_msg_pages(), a null page list pointer is handled the same as a message with 0-length trail, and both result in a "in_trail" variable set to false. The trail pointer is only used if in_trail is true. - The only other place the message trail pointer is used is out_msg_pos_next(). That function is only called by write_partial_msg_pages() and only touches the trail pointer if the in_trail value it is passed is true. Therefore a null ceph_msg->trail pointer is equivalent to a non-null pointer referring to a 0-length page list structure. Signed-off-by: Alex Elder <elder@inktank.com> Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
Diffstat (limited to 'net/ceph/osd_client.c')
-rw-r--r--net/ceph/osd_client.c43
1 files changed, 12 insertions, 31 deletions
diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c
index 7d38327a8e89..2be50d82ccbc 100644
--- a/net/ceph/osd_client.c
+++ b/net/ceph/osd_client.c
@@ -171,10 +171,7 @@ void ceph_osdc_release_request(struct kref *kref)
171 bio_put(req->r_bio); 171 bio_put(req->r_bio);
172#endif 172#endif
173 ceph_put_snap_context(req->r_snapc); 173 ceph_put_snap_context(req->r_snapc);
174 if (req->r_trail) { 174 ceph_pagelist_release(&req->r_trail);
175 ceph_pagelist_release(req->r_trail);
176 kfree(req->r_trail);
177 }
178 if (req->r_mempool) 175 if (req->r_mempool)
179 mempool_free(req, req->r_osdc->req_mempool); 176 mempool_free(req, req->r_osdc->req_mempool);
180 else 177 else
@@ -208,8 +205,7 @@ struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
208{ 205{
209 struct ceph_osd_request *req; 206 struct ceph_osd_request *req;
210 struct ceph_msg *msg; 207 struct ceph_msg *msg;
211 int needs_trail; 208 int num_op = get_num_ops(ops, NULL);
212 int num_op = get_num_ops(ops, &needs_trail);
213 size_t msg_size = sizeof(struct ceph_osd_request_head); 209 size_t msg_size = sizeof(struct ceph_osd_request_head);
214 210
215 msg_size += num_op*sizeof(struct ceph_osd_op); 211 msg_size += num_op*sizeof(struct ceph_osd_op);
@@ -252,15 +248,7 @@ struct ceph_osd_request *ceph_osdc_alloc_request(struct ceph_osd_client *osdc,
252 } 248 }
253 req->r_reply = msg; 249 req->r_reply = msg;
254 250
255 /* allocate space for the trailing data */ 251 ceph_pagelist_init(&req->r_trail);
256 if (needs_trail) {
257 req->r_trail = kmalloc(sizeof(struct ceph_pagelist), gfp_flags);
258 if (!req->r_trail) {
259 ceph_osdc_put_request(req);
260 return NULL;
261 }
262 ceph_pagelist_init(req->r_trail);
263 }
264 252
265 /* create request message; allow space for oid */ 253 /* create request message; allow space for oid */
266 msg_size += MAX_OBJ_NAME_SIZE; 254 msg_size += MAX_OBJ_NAME_SIZE;
@@ -312,29 +300,25 @@ static void osd_req_encode_op(struct ceph_osd_request *req,
312 case CEPH_OSD_OP_GETXATTR: 300 case CEPH_OSD_OP_GETXATTR:
313 case CEPH_OSD_OP_SETXATTR: 301 case CEPH_OSD_OP_SETXATTR:
314 case CEPH_OSD_OP_CMPXATTR: 302 case CEPH_OSD_OP_CMPXATTR:
315 BUG_ON(!req->r_trail);
316
317 dst->xattr.name_len = cpu_to_le32(src->xattr.name_len); 303 dst->xattr.name_len = cpu_to_le32(src->xattr.name_len);
318 dst->xattr.value_len = cpu_to_le32(src->xattr.value_len); 304 dst->xattr.value_len = cpu_to_le32(src->xattr.value_len);
319 dst->xattr.cmp_op = src->xattr.cmp_op; 305 dst->xattr.cmp_op = src->xattr.cmp_op;
320 dst->xattr.cmp_mode = src->xattr.cmp_mode; 306 dst->xattr.cmp_mode = src->xattr.cmp_mode;
321 ceph_pagelist_append(req->r_trail, src->xattr.name, 307 ceph_pagelist_append(&req->r_trail, src->xattr.name,
322 src->xattr.name_len); 308 src->xattr.name_len);
323 ceph_pagelist_append(req->r_trail, src->xattr.val, 309 ceph_pagelist_append(&req->r_trail, src->xattr.val,
324 src->xattr.value_len); 310 src->xattr.value_len);
325 break; 311 break;
326 case CEPH_OSD_OP_CALL: 312 case CEPH_OSD_OP_CALL:
327 BUG_ON(!req->r_trail);
328
329 dst->cls.class_len = src->cls.class_len; 313 dst->cls.class_len = src->cls.class_len;
330 dst->cls.method_len = src->cls.method_len; 314 dst->cls.method_len = src->cls.method_len;
331 dst->cls.indata_len = cpu_to_le32(src->cls.indata_len); 315 dst->cls.indata_len = cpu_to_le32(src->cls.indata_len);
332 316
333 ceph_pagelist_append(req->r_trail, src->cls.class_name, 317 ceph_pagelist_append(&req->r_trail, src->cls.class_name,
334 src->cls.class_len); 318 src->cls.class_len);
335 ceph_pagelist_append(req->r_trail, src->cls.method_name, 319 ceph_pagelist_append(&req->r_trail, src->cls.method_name,
336 src->cls.method_len); 320 src->cls.method_len);
337 ceph_pagelist_append(req->r_trail, src->cls.indata, 321 ceph_pagelist_append(&req->r_trail, src->cls.indata,
338 src->cls.indata_len); 322 src->cls.indata_len);
339 break; 323 break;
340 case CEPH_OSD_OP_ROLLBACK: 324 case CEPH_OSD_OP_ROLLBACK:
@@ -347,11 +331,9 @@ static void osd_req_encode_op(struct ceph_osd_request *req,
347 __le32 prot_ver = cpu_to_le32(src->watch.prot_ver); 331 __le32 prot_ver = cpu_to_le32(src->watch.prot_ver);
348 __le32 timeout = cpu_to_le32(src->watch.timeout); 332 __le32 timeout = cpu_to_le32(src->watch.timeout);
349 333
350 BUG_ON(!req->r_trail); 334 ceph_pagelist_append(&req->r_trail,
351
352 ceph_pagelist_append(req->r_trail,
353 &prot_ver, sizeof(prot_ver)); 335 &prot_ver, sizeof(prot_ver));
354 ceph_pagelist_append(req->r_trail, 336 ceph_pagelist_append(&req->r_trail,
355 &timeout, sizeof(timeout)); 337 &timeout, sizeof(timeout));
356 } 338 }
357 case CEPH_OSD_OP_NOTIFY_ACK: 339 case CEPH_OSD_OP_NOTIFY_ACK:
@@ -414,8 +396,7 @@ void ceph_osdc_build_request(struct ceph_osd_request *req,
414 op++; 396 op++;
415 } 397 }
416 398
417 if (req->r_trail) 399 data_len += req->r_trail.length;
418 data_len += req->r_trail->length;
419 400
420 if (snapc) { 401 if (snapc) {
421 head->snap_seq = cpu_to_le64(snapc->seq); 402 head->snap_seq = cpu_to_le64(snapc->seq);
@@ -1715,7 +1696,7 @@ int ceph_osdc_start_request(struct ceph_osd_client *osdc,
1715#ifdef CONFIG_BLOCK 1696#ifdef CONFIG_BLOCK
1716 req->r_request->bio = req->r_bio; 1697 req->r_request->bio = req->r_bio;
1717#endif 1698#endif
1718 req->r_request->trail = req->r_trail; 1699 req->r_request->trail = &req->r_trail;
1719 1700
1720 register_request(osdc, req); 1701 register_request(osdc, req);
1721 1702