root / xseg / peers / user / mt-mapperd.c @ d8a852fa
History | View | Annotate | Download (80 kB)
| 1 |
/*
|
|---|---|
| 2 |
* Copyright 2012 GRNET S.A. All rights reserved.
|
| 3 |
*
|
| 4 |
* Redistribution and use in source and binary forms, with or
|
| 5 |
* without modification, are permitted provided that the following
|
| 6 |
* conditions are met:
|
| 7 |
*
|
| 8 |
* 1. Redistributions of source code must retain the above
|
| 9 |
* copyright notice, this list of conditions and the following
|
| 10 |
* disclaimer.
|
| 11 |
* 2. Redistributions in binary form must reproduce the above
|
| 12 |
* copyright notice, this list of conditions and the following
|
| 13 |
* disclaimer in the documentation and/or other materials
|
| 14 |
* provided with the distribution.
|
| 15 |
*
|
| 16 |
* THIS SOFTWARE IS PROVIDED BY GRNET S.A. ``AS IS'' AND ANY EXPRESS
|
| 17 |
* OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
|
| 18 |
* WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
|
| 19 |
* PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL GRNET S.A OR
|
| 20 |
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
| 21 |
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
| 22 |
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
|
| 23 |
* USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED
|
| 24 |
* AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
| 25 |
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
|
| 26 |
* ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
|
| 27 |
* POSSIBILITY OF SUCH DAMAGE.
|
| 28 |
*
|
| 29 |
* The views and conclusions contained in the software and
|
| 30 |
* documentation are those of the authors and should not be
|
| 31 |
* interpreted as representing official policies, either expressed
|
| 32 |
* or implied, of GRNET S.A.
|
| 33 |
*/
|
| 34 |
|
| 35 |
#include <stdio.h> |
| 36 |
#include <unistd.h> |
| 37 |
#include <sys/types.h> |
| 38 |
#include <pthread.h> |
| 39 |
#include <xseg/xseg.h> |
| 40 |
#include <peer.h> |
| 41 |
#include <time.h> |
| 42 |
#include <xtypes/xlock.h> |
| 43 |
#include <xtypes/xhash.h> |
| 44 |
#include <xseg/protocol.h> |
| 45 |
#include <sys/stat.h> |
| 46 |
#include <fcntl.h> |
| 47 |
#include <errno.h> |
| 48 |
#include <sched.h> |
| 49 |
#include <sys/syscall.h> |
| 50 |
#include <openssl/sha.h> |
| 51 |
#include <ctype.h> |
| 52 |
|
| 53 |
/* general mapper flags */
|
| 54 |
#define MF_LOAD (1 << 0) |
| 55 |
#define MF_EXCLUSIVE (1 << 1) |
| 56 |
#define MF_FORCE (1 << 2) |
| 57 |
#define MF_ARCHIP (1 << 3) |
| 58 |
|
| 59 |
#ifndef SHA256_DIGEST_SIZE
|
| 60 |
#define SHA256_DIGEST_SIZE 32 |
| 61 |
#endif
|
| 62 |
/* hex representation of sha256 value takes up double the sha256 size */
|
| 63 |
#define HEXLIFIED_SHA256_DIGEST_SIZE (SHA256_DIGEST_SIZE << 1) |
| 64 |
|
| 65 |
#define block_size (1<<22) //FIXME this should be defined here? |
| 66 |
|
| 67 |
/* transparency byte + max object len in disk */
|
| 68 |
#define objectsize_in_map (1 + SHA256_DIGEST_SIZE) |
| 69 |
|
| 70 |
/* Map header contains:
|
| 71 |
* map version
|
| 72 |
* volume size
|
| 73 |
*/
|
| 74 |
#define mapheader_size (sizeof (uint32_t) + sizeof(uint64_t)) |
| 75 |
|
| 76 |
|
| 77 |
#define MAPPER_PREFIX "archip_" |
| 78 |
#define MAPPER_PREFIX_LEN 7 |
| 79 |
|
| 80 |
#define MAX_REAL_VOLUME_LEN (XSEG_MAX_TARGETLEN - MAPPER_PREFIX_LEN)
|
| 81 |
#define MAX_VOLUME_LEN (MAPPER_PREFIX_LEN + MAX_REAL_VOLUME_LEN)
|
| 82 |
|
| 83 |
#if MAX_VOLUME_LEN > XSEG_MAX_TARGETLEN
|
| 84 |
#error "XSEG_MAX_TARGETLEN should be at least MAX_VOLUME_LEN" |
| 85 |
#endif
|
| 86 |
|
| 87 |
#define MAX_OBJECT_LEN (MAPPER_PREFIX_LEN + HEXLIFIED_SHA256_DIGEST_SIZE)
|
| 88 |
|
| 89 |
#if MAX_OBJECT_LEN > XSEG_MAX_TARGETLEN
|
| 90 |
#error "XSEG_MAX_TARGETLEN should be at least MAX_OBJECT_LEN" |
| 91 |
#endif
|
| 92 |
|
| 93 |
#define MAX_VOLUME_SIZE \
|
| 94 |
((uint64_t) (((block_size-mapheader_size)/objectsize_in_map)* block_size)) |
| 95 |
|
| 96 |
|
| 97 |
//char *zero_block="0000000000000000000000000000000000000000000000000000000000000000";
|
| 98 |
|
| 99 |
/* pithos considers this a block full of zeros, so should we.
|
| 100 |
* it is actually the sha256 hash of nothing.
|
| 101 |
*/
|
| 102 |
char *zero_block="e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855"; |
| 103 |
#define ZERO_BLOCK_LEN (64) /* strlen(zero_block) */ |
| 104 |
|
| 105 |
/* dispatch_internal mapper states */
|
| 106 |
enum mapper_state {
|
| 107 |
ACCEPTED = 0,
|
| 108 |
WRITING = 1,
|
| 109 |
COPYING = 2,
|
| 110 |
DELETING = 3,
|
| 111 |
DROPPING_CACHE = 4
|
| 112 |
}; |
| 113 |
|
| 114 |
typedef void (*cb_t)(struct peer_req *pr, struct xseg_request *req); |
| 115 |
|
| 116 |
|
| 117 |
/* mapper object flags */
|
| 118 |
#define MF_OBJECT_EXIST (1 << 0) |
| 119 |
#define MF_OBJECT_COPYING (1 << 1) |
| 120 |
#define MF_OBJECT_WRITING (1 << 2) |
| 121 |
#define MF_OBJECT_DELETING (1 << 3) |
| 122 |
#define MF_OBJECT_DESTROYED (1 << 5) |
| 123 |
#define MF_OBJECT_SNAPSHOTTING (1 << 6) |
| 124 |
|
| 125 |
#define MF_OBJECT_NOT_READY (MF_OBJECT_COPYING|MF_OBJECT_WRITING|\
|
| 126 |
MF_OBJECT_DELETING|MF_OBJECT_SNAPSHOTTING) |
| 127 |
struct map_node {
|
| 128 |
uint32_t flags; |
| 129 |
uint32_t objectidx; |
| 130 |
uint32_t objectlen; |
| 131 |
char object[MAX_OBJECT_LEN + 1]; /* NULL terminated string */ |
| 132 |
struct map *map;
|
| 133 |
uint32_t ref; |
| 134 |
uint32_t waiters; |
| 135 |
st_cond_t cond; |
| 136 |
}; |
| 137 |
|
| 138 |
|
| 139 |
#define wait_on_pr(__pr, __condition__) \
|
| 140 |
while (__condition__){ \
|
| 141 |
ta--; \ |
| 142 |
__get_mapper_io(pr)->active = 0;\
|
| 143 |
XSEGLOG2(&lc, D, "Waiting on pr %lx, ta: %u", pr, ta); \
|
| 144 |
st_cond_wait(__pr->cond); \ |
| 145 |
} |
| 146 |
|
| 147 |
#define wait_on_mapnode(__mn, __condition__) \
|
| 148 |
while (__condition__){ \
|
| 149 |
ta--; \ |
| 150 |
__mn->waiters++; \ |
| 151 |
XSEGLOG2(&lc, D, "Waiting on map node %lx %s, waiters: %u, \
|
| 152 |
ta: %u", __mn, __mn->object, __mn->waiters, ta); \
|
| 153 |
st_cond_wait(__mn->cond); \ |
| 154 |
} |
| 155 |
|
| 156 |
#define wait_on_map(__map, __condition__) \
|
| 157 |
while (__condition__){ \
|
| 158 |
ta--; \ |
| 159 |
__map->waiters++; \ |
| 160 |
XSEGLOG2(&lc, D, "Waiting on map %lx %s, waiters: %u, ta: %u",\
|
| 161 |
__map, __map->volume, __map->waiters, ta); \ |
| 162 |
st_cond_wait(__map->cond); \ |
| 163 |
} |
| 164 |
|
| 165 |
#define signal_pr(__pr) \
|
| 166 |
do { \
|
| 167 |
if (!__get_mapper_io(pr)->active){\
|
| 168 |
ta++; \ |
| 169 |
XSEGLOG2(&lc, D, "Signaling pr %lx, ta: %u", pr, ta);\
|
| 170 |
__get_mapper_io(pr)->active = 1;\
|
| 171 |
st_cond_signal(__pr->cond); \ |
| 172 |
} \ |
| 173 |
}while(0) |
| 174 |
|
| 175 |
#define signal_map(__map) \
|
| 176 |
do { \
|
| 177 |
if (__map->waiters) { \
|
| 178 |
ta += 1; \
|
| 179 |
XSEGLOG2(&lc, D, "Signaling map %lx %s, waiters: %u, \
|
| 180 |
ta: %u", __map, __map->volume, __map->waiters, ta); \
|
| 181 |
__map->waiters--; \ |
| 182 |
st_cond_signal(__map->cond); \ |
| 183 |
} \ |
| 184 |
}while(0) |
| 185 |
|
| 186 |
#define signal_mapnode(__mn) \
|
| 187 |
do { \
|
| 188 |
if (__mn->waiters) { \
|
| 189 |
ta += __mn->waiters; \ |
| 190 |
XSEGLOG2(&lc, D, "Signaling map node %lx %s, waiters: \
|
| 191 |
%u, ta: %u", __mn, __mn->object, __mn->waiters, ta); \
|
| 192 |
__mn->waiters = 0; \
|
| 193 |
st_cond_broadcast(__mn->cond); \ |
| 194 |
} \ |
| 195 |
}while(0) |
| 196 |
|
| 197 |
|
| 198 |
/* map flags */
|
| 199 |
#define MF_MAP_LOADING (1 << 0) |
| 200 |
#define MF_MAP_DESTROYED (1 << 1) |
| 201 |
#define MF_MAP_WRITING (1 << 2) |
| 202 |
#define MF_MAP_DELETING (1 << 3) |
| 203 |
#define MF_MAP_DROPPING_CACHE (1 << 4) |
| 204 |
#define MF_MAP_EXCLUSIVE (1 << 5) |
| 205 |
#define MF_MAP_OPENING (1 << 6) |
| 206 |
#define MF_MAP_CLOSING (1 << 7) |
| 207 |
#define MF_MAP_DELETED (1 << 8) |
| 208 |
#define MF_MAP_SNAPSHOTTING (1 << 9) |
| 209 |
|
| 210 |
#define MF_MAP_NOT_READY (MF_MAP_LOADING|MF_MAP_WRITING|MF_MAP_DELETING|\
|
| 211 |
MF_MAP_DROPPING_CACHE|MF_MAP_OPENING| \ |
| 212 |
MF_MAP_SNAPSHOTTING) |
| 213 |
|
| 214 |
struct map {
|
| 215 |
uint32_t version; |
| 216 |
uint32_t flags; |
| 217 |
uint64_t size; |
| 218 |
uint32_t volumelen; |
| 219 |
char volume[MAX_VOLUME_LEN + 1]; /* NULL terminated string */ |
| 220 |
xhash_t *objects; /* obj_index --> map_node */
|
| 221 |
uint32_t ref; |
| 222 |
uint32_t waiters; |
| 223 |
st_cond_t cond; |
| 224 |
}; |
| 225 |
|
| 226 |
struct mapperd {
|
| 227 |
xport bportno; /* blocker that accesses data */
|
| 228 |
xport mbportno; /* blocker that accesses maps */
|
| 229 |
xhash_t *hashmaps; // hash_function(target) --> struct map
|
| 230 |
}; |
| 231 |
|
| 232 |
struct mapper_io {
|
| 233 |
volatile uint32_t copyups; /* nr of copyups pending, issued by this mapper io */ |
| 234 |
xhash_t *copyups_nodes; /* hash map (xseg_request) --> (corresponding map_node of copied up object)*/
|
| 235 |
struct map_node *copyup_node;
|
| 236 |
volatile int err; /* error flag */ |
| 237 |
volatile uint64_t del_pending;
|
| 238 |
volatile uint64_t snap_pending;
|
| 239 |
uint64_t delobj; |
| 240 |
uint64_t dcobj; |
| 241 |
cb_t cb; |
| 242 |
enum mapper_state state;
|
| 243 |
volatile int active; |
| 244 |
}; |
| 245 |
|
| 246 |
/* global vars */
|
| 247 |
struct mapperd *mapper;
|
| 248 |
|
| 249 |
void print_map(struct map *m); |
| 250 |
|
| 251 |
|
| 252 |
void custom_peer_usage()
|
| 253 |
{
|
| 254 |
fprintf(stderr, "Custom peer options: \n"
|
| 255 |
"-bp : port for block blocker(!)\n"
|
| 256 |
"-mbp : port for map blocker\n"
|
| 257 |
"\n");
|
| 258 |
} |
| 259 |
|
| 260 |
|
| 261 |
/*
|
| 262 |
* Helper functions
|
| 263 |
*/
|
| 264 |
|
| 265 |
static inline struct mapperd * __get_mapperd(struct peerd *peer) |
| 266 |
{
|
| 267 |
return (struct mapperd *) peer->priv; |
| 268 |
} |
| 269 |
|
| 270 |
static inline struct mapper_io * __get_mapper_io(struct peer_req *pr) |
| 271 |
{
|
| 272 |
return (struct mapper_io *) pr->priv; |
| 273 |
} |
| 274 |
|
| 275 |
static inline uint64_t calc_map_obj(struct map *map) |
| 276 |
{
|
| 277 |
if (map->size == -1) |
| 278 |
return 0; |
| 279 |
uint64_t nr_objs = map->size / block_size; |
| 280 |
if (map->size % block_size)
|
| 281 |
nr_objs++; |
| 282 |
return nr_objs;
|
| 283 |
} |
| 284 |
|
| 285 |
static uint32_t calc_nr_obj(struct xseg_request *req) |
| 286 |
{
|
| 287 |
unsigned int r = 1; |
| 288 |
uint64_t rem_size = req->size; |
| 289 |
uint64_t obj_offset = req->offset & (block_size -1); //modulo |
| 290 |
uint64_t obj_size = (rem_size + obj_offset > block_size) ? block_size - obj_offset : rem_size; |
| 291 |
rem_size -= obj_size; |
| 292 |
while (rem_size > 0) { |
| 293 |
obj_size = (rem_size > block_size) ? block_size : rem_size; |
| 294 |
rem_size -= obj_size; |
| 295 |
r++; |
| 296 |
} |
| 297 |
|
| 298 |
return r;
|
| 299 |
} |
| 300 |
|
| 301 |
/* hexlify function.
|
| 302 |
* Unsafe. Doesn't check if data length is odd!
|
| 303 |
*/
|
| 304 |
|
| 305 |
static void hexlify(unsigned char *data, char *hex) |
| 306 |
{
|
| 307 |
int i;
|
| 308 |
for (i=0; i<SHA256_DIGEST_LENGTH; i++) |
| 309 |
sprintf(hex+2*i, "%02x", data[i]); |
| 310 |
} |
| 311 |
|
| 312 |
static void unhexlify(char *hex, unsigned char *data) |
| 313 |
{
|
| 314 |
int i;
|
| 315 |
char c;
|
| 316 |
for (i=0; i<SHA256_DIGEST_LENGTH; i++){ |
| 317 |
data[i] = 0;
|
| 318 |
c = hex[2*i];
|
| 319 |
if (isxdigit(c)){
|
| 320 |
if (isdigit(c)){
|
| 321 |
c-= '0';
|
| 322 |
} |
| 323 |
else {
|
| 324 |
c = tolower(c); |
| 325 |
c = c-'a' + 10; |
| 326 |
} |
| 327 |
} |
| 328 |
else {
|
| 329 |
c = 0;
|
| 330 |
} |
| 331 |
data[i] |= (c << 4) & 0xF0; |
| 332 |
c = hex[2*i+1]; |
| 333 |
if (isxdigit(c)){
|
| 334 |
if (isdigit(c)){
|
| 335 |
c-= '0';
|
| 336 |
} |
| 337 |
else {
|
| 338 |
c = tolower(c); |
| 339 |
c = c-'a' + 10; |
| 340 |
} |
| 341 |
} |
| 342 |
else {
|
| 343 |
c = 0;
|
| 344 |
} |
| 345 |
data[i] |= c & 0x0F;
|
| 346 |
} |
| 347 |
} |
| 348 |
|
| 349 |
void merkle_hash(unsigned char *hashes, unsigned long len, |
| 350 |
unsigned char hash[SHA256_DIGEST_SIZE]) |
| 351 |
{
|
| 352 |
uint32_t i, l, s = 2;
|
| 353 |
uint32_t nr = len/SHA256_DIGEST_SIZE; |
| 354 |
unsigned char *buf; |
| 355 |
unsigned char tmp_hash[SHA256_DIGEST_SIZE]; |
| 356 |
|
| 357 |
if (!nr){
|
| 358 |
SHA256(hashes, 0, hash);
|
| 359 |
return;
|
| 360 |
} |
| 361 |
if (nr == 1){ |
| 362 |
memcpy(hash, hashes, SHA256_DIGEST_SIZE); |
| 363 |
return;
|
| 364 |
} |
| 365 |
while (s < nr)
|
| 366 |
s = s << 1;
|
| 367 |
buf = malloc(sizeof(unsigned char)* SHA256_DIGEST_SIZE * s); |
| 368 |
memcpy(buf, hashes, nr * SHA256_DIGEST_SIZE); |
| 369 |
memset(buf + nr * SHA256_DIGEST_SIZE, 0, (s - nr) * SHA256_DIGEST_SIZE);
|
| 370 |
for (l = s; l > 1; l = l/2) { |
| 371 |
for (i = 0; i < l; i += 2) { |
| 372 |
SHA256(buf + (i * SHA256_DIGEST_SIZE), |
| 373 |
2 * SHA256_DIGEST_SIZE, tmp_hash);
|
| 374 |
memcpy(buf + (i/2 * SHA256_DIGEST_SIZE),
|
| 375 |
tmp_hash, SHA256_DIGEST_SIZE); |
| 376 |
} |
| 377 |
} |
| 378 |
memcpy(hash, buf, SHA256_DIGEST_SIZE); |
| 379 |
} |
| 380 |
|
| 381 |
/*
|
| 382 |
* Maps handling functions
|
| 383 |
*/
|
| 384 |
|
| 385 |
static struct map * find_map(struct mapperd *mapper, char *volume) |
| 386 |
{
|
| 387 |
struct map *m = NULL; |
| 388 |
int r = xhash_lookup(mapper->hashmaps, (xhashidx) volume,
|
| 389 |
(xhashidx *) &m); |
| 390 |
if (r < 0) |
| 391 |
return NULL; |
| 392 |
return m;
|
| 393 |
} |
| 394 |
|
| 395 |
static struct map * find_map_len(struct mapperd *mapper, char *target, |
| 396 |
uint32_t targetlen, uint32_t flags) |
| 397 |
{
|
| 398 |
char buf[XSEG_MAX_TARGETLEN+1]; |
| 399 |
if (flags & MF_ARCHIP){
|
| 400 |
strncpy(buf, MAPPER_PREFIX, MAPPER_PREFIX_LEN); |
| 401 |
strncpy(buf + MAPPER_PREFIX_LEN, target, targetlen); |
| 402 |
buf[MAPPER_PREFIX_LEN + targetlen] = 0;
|
| 403 |
targetlen += MAPPER_PREFIX_LEN; |
| 404 |
} |
| 405 |
else {
|
| 406 |
strncpy(buf, target, targetlen); |
| 407 |
buf[targetlen] = 0;
|
| 408 |
} |
| 409 |
|
| 410 |
if (targetlen > MAX_VOLUME_LEN){
|
| 411 |
XSEGLOG2(&lc, E, "Namelen %u too long. Max: %d",
|
| 412 |
targetlen, MAX_VOLUME_LEN); |
| 413 |
return NULL; |
| 414 |
} |
| 415 |
|
| 416 |
XSEGLOG2(&lc, D, "looking up map %s, len %u",
|
| 417 |
buf, targetlen); |
| 418 |
return find_map(mapper, buf);
|
| 419 |
} |
| 420 |
|
| 421 |
|
| 422 |
static int insert_map(struct mapperd *mapper, struct map *map) |
| 423 |
{
|
| 424 |
int r = -1; |
| 425 |
|
| 426 |
if (find_map(mapper, map->volume)){
|
| 427 |
XSEGLOG2(&lc, W, "Map %s found in hash maps", map->volume);
|
| 428 |
goto out;
|
| 429 |
} |
| 430 |
|
| 431 |
XSEGLOG2(&lc, D, "Inserting map %s, len: %d (map: %lx)",
|
| 432 |
map->volume, strlen(map->volume), (unsigned long) map); |
| 433 |
r = xhash_insert(mapper->hashmaps, (xhashidx) map->volume, (xhashidx) map); |
| 434 |
while (r == -XHASH_ERESIZE) {
|
| 435 |
xhashidx shift = xhash_grow_size_shift(mapper->hashmaps); |
| 436 |
xhash_t *new_hashmap = xhash_resize(mapper->hashmaps, shift, 0, NULL); |
| 437 |
if (!new_hashmap){
|
| 438 |
XSEGLOG2(&lc, E, "Cannot grow mapper->hashmaps to sizeshift %llu",
|
| 439 |
(unsigned long long) shift); |
| 440 |
goto out;
|
| 441 |
} |
| 442 |
mapper->hashmaps = new_hashmap; |
| 443 |
r = xhash_insert(mapper->hashmaps, (xhashidx) map->volume, (xhashidx) map); |
| 444 |
} |
| 445 |
out:
|
| 446 |
return r;
|
| 447 |
} |
| 448 |
|
| 449 |
static int remove_map(struct mapperd *mapper, struct map *map) |
| 450 |
{
|
| 451 |
int r = -1; |
| 452 |
|
| 453 |
//assert no pending pr on map
|
| 454 |
|
| 455 |
r = xhash_delete(mapper->hashmaps, (xhashidx) map->volume); |
| 456 |
while (r == -XHASH_ERESIZE) {
|
| 457 |
xhashidx shift = xhash_shrink_size_shift(mapper->hashmaps); |
| 458 |
xhash_t *new_hashmap = xhash_resize(mapper->hashmaps, shift, 0, NULL); |
| 459 |
if (!new_hashmap){
|
| 460 |
XSEGLOG2(&lc, E, "Cannot shrink mapper->hashmaps to sizeshift %llu",
|
| 461 |
(unsigned long long) shift); |
| 462 |
goto out;
|
| 463 |
} |
| 464 |
mapper->hashmaps = new_hashmap; |
| 465 |
r = xhash_delete(mapper->hashmaps, (xhashidx) map->volume); |
| 466 |
} |
| 467 |
out:
|
| 468 |
return r;
|
| 469 |
} |
| 470 |
|
| 471 |
static struct xseg_request * __close_map(struct peer_req *pr, struct map *map) |
| 472 |
{
|
| 473 |
int r;
|
| 474 |
xport p; |
| 475 |
struct peerd *peer = pr->peer;
|
| 476 |
struct xseg_request *req;
|
| 477 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 478 |
void *dummy;
|
| 479 |
|
| 480 |
XSEGLOG2(&lc, I, "Closing map %s", map->volume);
|
| 481 |
|
| 482 |
req = xseg_get_request(peer->xseg, pr->portno, mapper->mbportno, X_ALLOC); |
| 483 |
if (!req){
|
| 484 |
XSEGLOG2(&lc, E, "Cannot allocate request for map %s",
|
| 485 |
map->volume); |
| 486 |
goto out_err;
|
| 487 |
} |
| 488 |
|
| 489 |
r = xseg_prep_request(peer->xseg, req, map->volumelen, 0);
|
| 490 |
if (r < 0){ |
| 491 |
XSEGLOG2(&lc, E, "Cannot prepare request for map %s",
|
| 492 |
map->volume); |
| 493 |
goto out_put;
|
| 494 |
} |
| 495 |
|
| 496 |
char *reqtarget = xseg_get_target(peer->xseg, req);
|
| 497 |
if (!reqtarget)
|
| 498 |
goto out_put;
|
| 499 |
strncpy(reqtarget, map->volume, req->targetlen); |
| 500 |
req->op = X_RELEASE; |
| 501 |
req->size = 0;
|
| 502 |
req->offset = 0;
|
| 503 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 504 |
if (r < 0){ |
| 505 |
XSEGLOG2(&lc, E, "Cannot set request data for map %s",
|
| 506 |
map->volume); |
| 507 |
goto out_put;
|
| 508 |
} |
| 509 |
p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 510 |
if (p == NoPort){
|
| 511 |
XSEGLOG2(&lc, E, "Cannot submit request for map %s",
|
| 512 |
map->volume); |
| 513 |
goto out_unset;
|
| 514 |
} |
| 515 |
r = xseg_signal(peer->xseg, p); |
| 516 |
map->flags |= MF_MAP_CLOSING; |
| 517 |
|
| 518 |
XSEGLOG2(&lc, I, "Map %s closing", map->volume);
|
| 519 |
return req;
|
| 520 |
|
| 521 |
out_unset:
|
| 522 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 523 |
out_put:
|
| 524 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 525 |
out_err:
|
| 526 |
return NULL; |
| 527 |
} |
| 528 |
|
| 529 |
static int close_map(struct peer_req *pr, struct map *map) |
| 530 |
{
|
| 531 |
int err;
|
| 532 |
struct xseg_request *req;
|
| 533 |
struct peerd *peer = pr->peer;
|
| 534 |
|
| 535 |
req = __close_map(pr, map); |
| 536 |
if (!req)
|
| 537 |
return -1; |
| 538 |
wait_on_pr(pr, (!((req->state & XS_FAILED)||(req->state & XS_SERVED)))); |
| 539 |
map->flags &= ~MF_MAP_CLOSING; |
| 540 |
err = req->state & XS_FAILED; |
| 541 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 542 |
if (err)
|
| 543 |
return -1; |
| 544 |
return 0; |
| 545 |
} |
| 546 |
|
| 547 |
/*
|
| 548 |
static int find_or_load_map(struct peerd *peer, struct peer_req *pr,
|
| 549 |
char *target, uint32_t targetlen, struct map **m)
|
| 550 |
{
|
| 551 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 552 |
int r;
|
| 553 |
*m = find_map(mapper, target, targetlen);
|
| 554 |
if (*m) {
|
| 555 |
XSEGLOG2(&lc, D, "Found map %s (%u)", (*m)->volume, (unsigned long) *m);
|
| 556 |
if ((*m)->flags & MF_MAP_NOT_READY) {
|
| 557 |
__xq_append_tail(&(*m)->pending, (xqindex) pr);
|
| 558 |
XSEGLOG2(&lc, I, "Map %s found and not ready", (*m)->volume);
|
| 559 |
return MF_PENDING;
|
| 560 |
//} else if ((*m)->flags & MF_MAP_DESTROYED){
|
| 561 |
// return -1;
|
| 562 |
//
|
| 563 |
}else {
|
| 564 |
XSEGLOG2(&lc, I, "Map %s found", (*m)->volume);
|
| 565 |
return 0;
|
| 566 |
}
|
| 567 |
}
|
| 568 |
r = open_map(peer, pr, target, targetlen, 0);
|
| 569 |
if (r < 0)
|
| 570 |
return -1; //error
|
| 571 |
return MF_PENDING;
|
| 572 |
}
|
| 573 |
*/
|
| 574 |
/*
|
| 575 |
* Object handling functions
|
| 576 |
*/
|
| 577 |
|
| 578 |
struct map_node *find_object(struct map *map, uint64_t obj_index) |
| 579 |
{
|
| 580 |
struct map_node *mn;
|
| 581 |
int r = xhash_lookup(map->objects, obj_index, (xhashidx *) &mn);
|
| 582 |
if (r < 0) |
| 583 |
return NULL; |
| 584 |
return mn;
|
| 585 |
} |
| 586 |
|
| 587 |
static int insert_object(struct map *map, struct map_node *mn) |
| 588 |
{
|
| 589 |
//FIXME no find object first
|
| 590 |
int r = xhash_insert(map->objects, mn->objectidx, (xhashidx) mn);
|
| 591 |
if (r == -XHASH_ERESIZE) {
|
| 592 |
unsigned long shift = xhash_grow_size_shift(map->objects); |
| 593 |
map->objects = xhash_resize(map->objects, shift, 0, NULL); |
| 594 |
if (!map->objects)
|
| 595 |
return -1; |
| 596 |
r = xhash_insert(map->objects, mn->objectidx, (xhashidx) mn); |
| 597 |
} |
| 598 |
return r;
|
| 599 |
} |
| 600 |
|
| 601 |
|
| 602 |
/*
|
| 603 |
* map read/write functions
|
| 604 |
*
|
| 605 |
* version 0 -> pithos map
|
| 606 |
* version 1 -> archipelago version 1
|
| 607 |
*
|
| 608 |
*
|
| 609 |
* functions
|
| 610 |
* int read_object(struct map_node *mn, unsigned char *buf)
|
| 611 |
* int prepare_write_object(struct peer_req *pr, struct map *map,
|
| 612 |
* struct map_node *mn, struct xseg_request *req)
|
| 613 |
* int read_map(struct map *m, unsigned char * data)
|
| 614 |
* int prepare_write_map(struct peer_req *pr, struct map *map,
|
| 615 |
* struct xseg_request *req)
|
| 616 |
*/
|
| 617 |
|
| 618 |
struct map_functions {
|
| 619 |
int (*read_object)(struct map_node *mn, unsigned char *buf); |
| 620 |
int (*prepare_write_object)(struct peer_req *pr, struct map *map, |
| 621 |
struct map_node *mn, struct xseg_request *req); |
| 622 |
int (*read_map)(struct map *m, unsigned char * data); |
| 623 |
int (*prepare_write_map)(struct peer_req *pr, struct map *map, |
| 624 |
struct xseg_request *req);
|
| 625 |
}; |
| 626 |
|
| 627 |
/* version 0 functions */
|
| 628 |
|
| 629 |
/* no header */
|
| 630 |
#define v0_mapheader_size 0 |
| 631 |
/* just the unhexlified name */
|
| 632 |
#define v0_objectsize_in_map SHA256_DIGEST_SIZE
|
| 633 |
|
| 634 |
static inline int read_object_v0(struct map_node *mn, unsigned char *buf) |
| 635 |
{
|
| 636 |
hexlify(buf, mn->object); |
| 637 |
mn->object[HEXLIFIED_SHA256_DIGEST_SIZE] = 0;
|
| 638 |
mn->objectlen = HEXLIFIED_SHA256_DIGEST_SIZE; |
| 639 |
mn->flags = MF_OBJECT_EXIST; |
| 640 |
|
| 641 |
return 0; |
| 642 |
} |
| 643 |
|
| 644 |
static void v0_object_to_map(struct map_node *mn, unsigned char *data) |
| 645 |
{
|
| 646 |
unhexlify(mn->object, data); |
| 647 |
} |
| 648 |
|
| 649 |
static int prepare_write_object_v0(struct peer_req *pr, struct map *map, |
| 650 |
struct map_node *mn, struct xseg_request *req) |
| 651 |
{
|
| 652 |
struct peerd *peer = pr->peer;
|
| 653 |
int r = xseg_prep_request(peer->xseg, req, map->volumelen, v0_objectsize_in_map);
|
| 654 |
if (r < 0){ |
| 655 |
XSEGLOG2(&lc, E, "Cannot allocate request for object %s. \n\t"
|
| 656 |
"(Map: %s [%llu]",
|
| 657 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 658 |
return -1; |
| 659 |
} |
| 660 |
char *target = xseg_get_target(peer->xseg, req);
|
| 661 |
strncpy(target, map->volume, req->targetlen); |
| 662 |
req->size = req->datalen; |
| 663 |
req->offset = v0_mapheader_size + mn->objectidx * v0_objectsize_in_map; |
| 664 |
|
| 665 |
unsigned char *data = xseg_get_data(pr->peer->xseg, req); |
| 666 |
v0_object_to_map(mn, data); |
| 667 |
return -1; |
| 668 |
} |
| 669 |
|
| 670 |
static int read_map_v0(struct map *m, unsigned char * data) |
| 671 |
{
|
| 672 |
int r;
|
| 673 |
struct map_node *map_node;
|
| 674 |
uint64_t i; |
| 675 |
uint64_t pos = 0;
|
| 676 |
uint64_t max_nr_objs = block_size/SHA256_DIGEST_SIZE; |
| 677 |
XSEGLOG2(&lc, D, "Max nr_objs %llu", max_nr_objs);
|
| 678 |
char nulls[SHA256_DIGEST_SIZE];
|
| 679 |
memset(nulls, 0, SHA256_DIGEST_SIZE);
|
| 680 |
map_node = calloc(max_nr_objs, sizeof(struct map_node)); |
| 681 |
if (!map_node)
|
| 682 |
return -1; |
| 683 |
for (i = 0; i < max_nr_objs; i++) { |
| 684 |
if (!memcmp(data+pos, nulls, v0_objectsize_in_map))
|
| 685 |
break;
|
| 686 |
map_node[i].objectidx = i; |
| 687 |
map_node[i].map = m; |
| 688 |
map_node[i].waiters = 0;
|
| 689 |
map_node[i].ref = 1;
|
| 690 |
map_node[i].cond = st_cond_new(); //FIXME err check;
|
| 691 |
read_object_v0(&map_node[i], data+pos); |
| 692 |
pos += v0_objectsize_in_map; |
| 693 |
r = insert_object(m, &map_node[i]); //FIXME error check
|
| 694 |
} |
| 695 |
XSEGLOG2(&lc, D, "Found %llu objects", i);
|
| 696 |
m->size = i * block_size; |
| 697 |
return 0; |
| 698 |
} |
| 699 |
|
| 700 |
static int prepare_write_map_v0(struct peer_req *pr, struct map *map, |
| 701 |
struct xseg_request *req)
|
| 702 |
{
|
| 703 |
struct peerd *peer = pr->peer;
|
| 704 |
uint64_t i, pos = 0, max_objidx = calc_map_obj(map);
|
| 705 |
struct map_node *mn;
|
| 706 |
int r = xseg_prep_request(peer->xseg, req, map->volumelen,
|
| 707 |
v0_mapheader_size + max_objidx * v0_objectsize_in_map); |
| 708 |
if (r < 0){ |
| 709 |
XSEGLOG2(&lc, E, "Cannot prepare request for map %s", map->volume);
|
| 710 |
return -1; |
| 711 |
} |
| 712 |
char *target = xseg_get_target(peer->xseg, req);
|
| 713 |
strncpy(target, map->volume, req->targetlen); |
| 714 |
char *data = xseg_get_data(peer->xseg, req);
|
| 715 |
|
| 716 |
req->op = X_WRITE; |
| 717 |
req->size = req->datalen; |
| 718 |
req->offset = 0;
|
| 719 |
|
| 720 |
for (i = 0; i < max_objidx; i++) { |
| 721 |
mn = find_object(map, i); |
| 722 |
if (!mn){
|
| 723 |
XSEGLOG2(&lc, E, "Cannot find object %llu for map %s",
|
| 724 |
(unsigned long long) i, map->volume); |
| 725 |
return -1; |
| 726 |
} |
| 727 |
v0_object_to_map(mn, (unsigned char *)(data+pos)); |
| 728 |
pos += v0_objectsize_in_map; |
| 729 |
} |
| 730 |
XSEGLOG2(&lc, D, "Prepared %llu objects", i);
|
| 731 |
return 0; |
| 732 |
} |
| 733 |
|
| 734 |
/* static struct map_functions map_functions_v0 = { */
|
| 735 |
/* .read_object = read_object_v0, */
|
| 736 |
/* .read_map = read_map_v0, */
|
| 737 |
/* .prepare_write_object = prepare_write_object_v0, */
|
| 738 |
/* .prepare_write_map = prepare_write_map_v0 */
|
| 739 |
/* }; */
|
| 740 |
#define map_functions_v0 { \
|
| 741 |
.read_object = read_object_v0, \ |
| 742 |
.read_map = read_map_v0, \ |
| 743 |
.prepare_write_object = prepare_write_object_v0,\ |
| 744 |
.prepare_write_map = prepare_write_map_v0 \ |
| 745 |
} |
| 746 |
/* v1 functions */
|
| 747 |
|
| 748 |
/* transparency byte + max object len in disk */
|
| 749 |
#define v1_objectsize_in_map (1 + SHA256_DIGEST_SIZE) |
| 750 |
|
| 751 |
/* Map header contains:
|
| 752 |
* map version
|
| 753 |
* volume size
|
| 754 |
*/
|
| 755 |
#define v1_mapheader_size (sizeof (uint32_t) + sizeof(uint64_t)) |
| 756 |
|
| 757 |
static inline int read_object_v1(struct map_node *mn, unsigned char *buf) |
| 758 |
{
|
| 759 |
char c = buf[0]; |
| 760 |
mn->flags = 0;
|
| 761 |
if (c){
|
| 762 |
mn->flags |= MF_OBJECT_EXIST; |
| 763 |
strcpy(mn->object, MAPPER_PREFIX); |
| 764 |
hexlify(buf+1, mn->object + MAPPER_PREFIX_LEN);
|
| 765 |
mn->object[MAX_OBJECT_LEN] = 0;
|
| 766 |
mn->objectlen = strlen(mn->object); |
| 767 |
} |
| 768 |
else {
|
| 769 |
mn->flags &= ~MF_OBJECT_EXIST; |
| 770 |
hexlify(buf+1, mn->object);
|
| 771 |
mn->object[HEXLIFIED_SHA256_DIGEST_SIZE] = 0;
|
| 772 |
mn->objectlen = strlen(mn->object); |
| 773 |
} |
| 774 |
return 0; |
| 775 |
} |
| 776 |
|
| 777 |
static inline void v1_object_to_map(char* buf, struct map_node *mn) |
| 778 |
{
|
| 779 |
buf[0] = (mn->flags & MF_OBJECT_EXIST)? 1 : 0; |
| 780 |
if (buf[0]){ |
| 781 |
/* strip common prefix */
|
| 782 |
unhexlify(mn->object+MAPPER_PREFIX_LEN, (unsigned char *)(buf+1)); |
| 783 |
} |
| 784 |
else {
|
| 785 |
unhexlify(mn->object, (unsigned char *)(buf+1)); |
| 786 |
} |
| 787 |
} |
| 788 |
|
| 789 |
static int prepare_write_object_v1(struct peer_req *pr, struct map *map, |
| 790 |
struct map_node *mn, struct xseg_request *req) |
| 791 |
{
|
| 792 |
struct peerd *peer = pr->peer;
|
| 793 |
int r = xseg_prep_request(peer->xseg, req, map->volumelen, v1_objectsize_in_map);
|
| 794 |
if (r < 0){ |
| 795 |
XSEGLOG2(&lc, E, "Cannot allocate request for object %s. \n\t"
|
| 796 |
"(Map: %s [%llu]",
|
| 797 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 798 |
return -1; |
| 799 |
} |
| 800 |
char *target = xseg_get_target(peer->xseg, req);
|
| 801 |
strncpy(target, map->volume, req->targetlen); |
| 802 |
req->size = req->datalen; |
| 803 |
req->offset = v1_mapheader_size + mn->objectidx * v1_objectsize_in_map; |
| 804 |
|
| 805 |
char *data = xseg_get_data(pr->peer->xseg, req);
|
| 806 |
v1_object_to_map(data, mn); |
| 807 |
return 0; |
| 808 |
} |
| 809 |
|
| 810 |
static int read_map_v1(struct map *m, unsigned char * data) |
| 811 |
{
|
| 812 |
int r;
|
| 813 |
struct map_node *map_node;
|
| 814 |
uint64_t i; |
| 815 |
uint64_t pos = 0;
|
| 816 |
uint64_t nr_objs; |
| 817 |
|
| 818 |
/* read header */
|
| 819 |
m->version = *(uint32_t *) (data + pos); |
| 820 |
pos += sizeof(uint32_t);
|
| 821 |
m->size = *(uint64_t *) (data + pos); |
| 822 |
pos += sizeof(uint64_t);
|
| 823 |
|
| 824 |
/* read objects */
|
| 825 |
nr_objs = m->size / block_size; |
| 826 |
if (m->size % block_size)
|
| 827 |
nr_objs++; |
| 828 |
map_node = calloc(nr_objs, sizeof(struct map_node)); |
| 829 |
if (!map_node)
|
| 830 |
return -1; |
| 831 |
|
| 832 |
for (i = 0; i < nr_objs; i++) { |
| 833 |
map_node[i].map = m; |
| 834 |
map_node[i].objectidx = i; |
| 835 |
map_node[i].waiters = 0;
|
| 836 |
map_node[i].ref = 1;
|
| 837 |
map_node[i].cond = st_cond_new(); //FIXME err check;
|
| 838 |
read_object_v1(&map_node[i], data+pos); |
| 839 |
pos += objectsize_in_map; |
| 840 |
r = insert_object(m, &map_node[i]); //FIXME error check
|
| 841 |
} |
| 842 |
return 0; |
| 843 |
} |
| 844 |
|
| 845 |
static int prepare_write_map_v1(struct peer_req *pr, struct map *m, |
| 846 |
struct xseg_request *req)
|
| 847 |
{
|
| 848 |
struct peerd *peer = pr->peer;
|
| 849 |
uint64_t i, pos = 0, max_objidx = calc_map_obj(m);
|
| 850 |
struct map_node *mn;
|
| 851 |
|
| 852 |
int r = xseg_prep_request(peer->xseg, req, m->volumelen,
|
| 853 |
v1_mapheader_size + max_objidx * v1_objectsize_in_map); |
| 854 |
if (r < 0){ |
| 855 |
XSEGLOG2(&lc, E, "Cannot prepare request for map %s", m->volume);
|
| 856 |
return -1; |
| 857 |
} |
| 858 |
char *target = xseg_get_target(peer->xseg, req);
|
| 859 |
strncpy(target, m->volume, req->targetlen); |
| 860 |
char *data = xseg_get_data(peer->xseg, req);
|
| 861 |
|
| 862 |
memcpy(data + pos, &m->version, sizeof(m->version));
|
| 863 |
pos += sizeof(m->version);
|
| 864 |
memcpy(data + pos, &m->size, sizeof(m->size));
|
| 865 |
pos += sizeof(m->size);
|
| 866 |
|
| 867 |
req->op = X_WRITE; |
| 868 |
req->size = req->datalen; |
| 869 |
req->offset = 0;
|
| 870 |
|
| 871 |
for (i = 0; i < max_objidx; i++) { |
| 872 |
mn = find_object(m, i); |
| 873 |
if (!mn){
|
| 874 |
XSEGLOG2(&lc, E, "Cannot find object %lli for map %s",
|
| 875 |
(unsigned long long) i, m->volume); |
| 876 |
return -1; |
| 877 |
} |
| 878 |
v1_object_to_map(data+pos, mn); |
| 879 |
pos += v1_objectsize_in_map; |
| 880 |
} |
| 881 |
return 0; |
| 882 |
} |
| 883 |
|
| 884 |
/* static struct map_functions map_functions_v1 = { */
|
| 885 |
/* .read_object = read_object_v1, */
|
| 886 |
/* .read_map = read_map_v1, */
|
| 887 |
/* .prepare_write_object = prepare_write_object_v1, */
|
| 888 |
/* .prepare_write_map = prepare_write_map_v1 */
|
| 889 |
/* }; */
|
| 890 |
#define map_functions_v1 { \
|
| 891 |
.read_object = read_object_v1, \ |
| 892 |
.read_map = read_map_v1, \ |
| 893 |
.prepare_write_object = prepare_write_object_v1,\ |
| 894 |
.prepare_write_map = prepare_write_map_v1 \ |
| 895 |
} |
| 896 |
|
| 897 |
static struct map_functions map_functions[] = { map_functions_v0, |
| 898 |
map_functions_v1 }; |
| 899 |
#define MAP_LATEST_VERSION 1 |
| 900 |
/* end of functions */
|
| 901 |
|
| 902 |
|
| 903 |
|
| 904 |
|
| 905 |
|
| 906 |
static inline void pithosmap_to_object(struct map_node *mn, unsigned char *buf) |
| 907 |
{
|
| 908 |
hexlify(buf, mn->object); |
| 909 |
mn->object[HEXLIFIED_SHA256_DIGEST_SIZE] = 0;
|
| 910 |
mn->objectlen = HEXLIFIED_SHA256_DIGEST_SIZE; |
| 911 |
mn->flags = MF_OBJECT_EXIST; |
| 912 |
} |
| 913 |
|
| 914 |
static inline void map_to_object(struct map_node *mn, unsigned char *buf) |
| 915 |
{
|
| 916 |
char c = buf[0]; |
| 917 |
mn->flags = 0;
|
| 918 |
if (c){
|
| 919 |
mn->flags |= MF_OBJECT_EXIST; |
| 920 |
strcpy(mn->object, MAPPER_PREFIX); |
| 921 |
hexlify(buf+1, mn->object + MAPPER_PREFIX_LEN);
|
| 922 |
mn->object[MAX_OBJECT_LEN] = 0;
|
| 923 |
mn->objectlen = strlen(mn->object); |
| 924 |
} |
| 925 |
else {
|
| 926 |
hexlify(buf+1, mn->object);
|
| 927 |
mn->object[HEXLIFIED_SHA256_DIGEST_SIZE] = 0;
|
| 928 |
mn->objectlen = strlen(mn->object); |
| 929 |
} |
| 930 |
|
| 931 |
} |
| 932 |
|
| 933 |
static inline void object_to_map(char* buf, struct map_node *mn) |
| 934 |
{
|
| 935 |
buf[0] = (mn->flags & MF_OBJECT_EXIST)? 1 : 0; |
| 936 |
if (buf[0]){ |
| 937 |
/* strip common prefix */
|
| 938 |
unhexlify(mn->object+MAPPER_PREFIX_LEN, (unsigned char *)(buf+1)); |
| 939 |
} |
| 940 |
else {
|
| 941 |
unhexlify(mn->object, (unsigned char *)(buf+1)); |
| 942 |
} |
| 943 |
} |
| 944 |
|
| 945 |
static inline void mapheader_to_map(struct map *m, char *buf) |
| 946 |
{
|
| 947 |
uint64_t pos = 0;
|
| 948 |
memcpy(buf + pos, &m->version, sizeof(m->version));
|
| 949 |
pos += sizeof(m->version);
|
| 950 |
memcpy(buf + pos, &m->size, sizeof(m->size));
|
| 951 |
pos += sizeof(m->size);
|
| 952 |
} |
| 953 |
|
| 954 |
|
| 955 |
static struct xseg_request * object_write(struct peerd *peer, struct peer_req *pr, |
| 956 |
struct map *map, struct map_node *mn) |
| 957 |
{
|
| 958 |
int r;
|
| 959 |
void *dummy;
|
| 960 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 961 |
struct xseg_request *req = xseg_get_request(peer->xseg, pr->portno,
|
| 962 |
mapper->mbportno, X_ALLOC); |
| 963 |
if (!req){
|
| 964 |
XSEGLOG2(&lc, E, "Cannot allocate request for object %s. \n\t"
|
| 965 |
"(Map: %s [%llu]",
|
| 966 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 967 |
goto out_err;
|
| 968 |
} |
| 969 |
|
| 970 |
r = map_functions[map->version].prepare_write_object(pr, map, mn, req); |
| 971 |
if (r < 0){ |
| 972 |
XSEGLOG2(&lc, E, "Cannot prepare write object");
|
| 973 |
goto out_put;
|
| 974 |
} |
| 975 |
req->op = X_WRITE; |
| 976 |
|
| 977 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 978 |
if (r < 0){ |
| 979 |
XSEGLOG2(&lc, E, "Cannot set request data for object %s. \n\t"
|
| 980 |
"(Map: %s [%llu]",
|
| 981 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 982 |
goto out_put;
|
| 983 |
} |
| 984 |
xport p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 985 |
if (p == NoPort){
|
| 986 |
XSEGLOG2(&lc, E, "Cannot submit request for object %s. \n\t"
|
| 987 |
"(Map: %s [%llu]",
|
| 988 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 989 |
goto out_unset;
|
| 990 |
} |
| 991 |
r = xseg_signal(peer->xseg, p); |
| 992 |
if (r < 0) |
| 993 |
XSEGLOG2(&lc, W, "Cannot signal port %u", p);
|
| 994 |
|
| 995 |
XSEGLOG2(&lc, I, "Writing object %s \n\t"
|
| 996 |
"Map: %s [%llu]",
|
| 997 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 998 |
|
| 999 |
return req;
|
| 1000 |
|
| 1001 |
out_unset:
|
| 1002 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 1003 |
out_put:
|
| 1004 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1005 |
out_err:
|
| 1006 |
XSEGLOG2(&lc, E, "Object write for object %s failed. \n\t"
|
| 1007 |
"(Map: %s [%llu]",
|
| 1008 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 1009 |
return NULL; |
| 1010 |
} |
| 1011 |
|
| 1012 |
static struct xseg_request * __write_map(struct peer_req* pr, struct map *map) |
| 1013 |
{
|
| 1014 |
int r;
|
| 1015 |
void *dummy;
|
| 1016 |
struct peerd *peer = pr->peer;
|
| 1017 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1018 |
struct xseg_request *req = xseg_get_request(peer->xseg, pr->portno,
|
| 1019 |
mapper->mbportno, X_ALLOC); |
| 1020 |
if (!req){
|
| 1021 |
XSEGLOG2(&lc, E, "Cannot allocate request for map %s", map->volume);
|
| 1022 |
goto out_err;
|
| 1023 |
} |
| 1024 |
|
| 1025 |
r = map_functions[map->version].prepare_write_map(pr, map, req); |
| 1026 |
if (r < 0){ |
| 1027 |
XSEGLOG2(&lc, E, "Cannot prepare write map");
|
| 1028 |
goto out_put;
|
| 1029 |
} |
| 1030 |
|
| 1031 |
req->op = X_WRITE; |
| 1032 |
|
| 1033 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 1034 |
if (r < 0){ |
| 1035 |
XSEGLOG2(&lc, E, "Cannot set request data for map %s",
|
| 1036 |
map->volume); |
| 1037 |
goto out_put;
|
| 1038 |
} |
| 1039 |
xport p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 1040 |
if (p == NoPort){
|
| 1041 |
XSEGLOG2(&lc, E, "Cannot submit request for map %s",
|
| 1042 |
map->volume); |
| 1043 |
goto out_unset;
|
| 1044 |
} |
| 1045 |
r = xseg_signal(peer->xseg, p); |
| 1046 |
if (r < 0) |
| 1047 |
XSEGLOG2(&lc, W, "Cannot signal port %u", p);
|
| 1048 |
|
| 1049 |
map->flags |= MF_MAP_WRITING; |
| 1050 |
XSEGLOG2(&lc, I, "Writing map %s", map->volume);
|
| 1051 |
return req;
|
| 1052 |
|
| 1053 |
out_unset:
|
| 1054 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 1055 |
out_put:
|
| 1056 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1057 |
out_err:
|
| 1058 |
XSEGLOG2(&lc, E, "Map write for map %s failed.", map->volume);
|
| 1059 |
return NULL; |
| 1060 |
} |
| 1061 |
|
| 1062 |
static int write_map(struct peer_req* pr, struct map *map) |
| 1063 |
{
|
| 1064 |
int r = 0; |
| 1065 |
struct peerd *peer = pr->peer;
|
| 1066 |
struct xseg_request *req = __write_map(pr, map);
|
| 1067 |
if (!req)
|
| 1068 |
return -1; |
| 1069 |
wait_on_pr(pr, (!(req->state & XS_FAILED || req->state & XS_SERVED))); |
| 1070 |
if (req->state & XS_FAILED)
|
| 1071 |
r = -1;
|
| 1072 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1073 |
map->flags &= ~MF_MAP_WRITING; |
| 1074 |
return r;
|
| 1075 |
} |
| 1076 |
|
| 1077 |
static struct xseg_request * __load_map(struct peer_req *pr, struct map *m) |
| 1078 |
{
|
| 1079 |
int r;
|
| 1080 |
xport p; |
| 1081 |
struct xseg_request *req;
|
| 1082 |
struct peerd *peer = pr->peer;
|
| 1083 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1084 |
void *dummy;
|
| 1085 |
|
| 1086 |
XSEGLOG2(&lc, I, "Loading ng map %s", m->volume);
|
| 1087 |
|
| 1088 |
req = xseg_get_request(peer->xseg, pr->portno, mapper->mbportno, X_ALLOC); |
| 1089 |
if (!req){
|
| 1090 |
XSEGLOG2(&lc, E, "Cannot allocate request for map %s",
|
| 1091 |
m->volume); |
| 1092 |
goto out_fail;
|
| 1093 |
} |
| 1094 |
|
| 1095 |
r = xseg_prep_request(peer->xseg, req, m->volumelen, block_size); |
| 1096 |
if (r < 0){ |
| 1097 |
XSEGLOG2(&lc, E, "Cannot prepare request for map %s",
|
| 1098 |
m->volume); |
| 1099 |
goto out_put;
|
| 1100 |
} |
| 1101 |
|
| 1102 |
char *reqtarget = xseg_get_target(peer->xseg, req);
|
| 1103 |
if (!reqtarget)
|
| 1104 |
goto out_put;
|
| 1105 |
strncpy(reqtarget, m->volume, req->targetlen); |
| 1106 |
req->op = X_READ; |
| 1107 |
req->size = block_size; |
| 1108 |
req->offset = 0;
|
| 1109 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 1110 |
if (r < 0){ |
| 1111 |
XSEGLOG2(&lc, E, "Cannot set request data for map %s",
|
| 1112 |
m->volume); |
| 1113 |
goto out_put;
|
| 1114 |
} |
| 1115 |
p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 1116 |
if (p == NoPort){
|
| 1117 |
XSEGLOG2(&lc, E, "Cannot submit request for map %s",
|
| 1118 |
m->volume); |
| 1119 |
goto out_unset;
|
| 1120 |
} |
| 1121 |
r = xseg_signal(peer->xseg, p); |
| 1122 |
|
| 1123 |
m->flags |= MF_MAP_LOADING; |
| 1124 |
XSEGLOG2(&lc, I, "Map %s loading", m->volume);
|
| 1125 |
return req;
|
| 1126 |
|
| 1127 |
out_unset:
|
| 1128 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 1129 |
out_put:
|
| 1130 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1131 |
out_fail:
|
| 1132 |
return NULL; |
| 1133 |
} |
| 1134 |
|
| 1135 |
static int read_map (struct map *map, unsigned char *buf) |
| 1136 |
{
|
| 1137 |
char nulls[SHA256_DIGEST_SIZE];
|
| 1138 |
memset(nulls, 0, SHA256_DIGEST_SIZE);
|
| 1139 |
|
| 1140 |
int r = !memcmp(buf, nulls, SHA256_DIGEST_SIZE);
|
| 1141 |
if (r) {
|
| 1142 |
XSEGLOG2(&lc, E, "Read zeros");
|
| 1143 |
return -1; |
| 1144 |
} |
| 1145 |
//type 1, archip type, type 0 pithos map
|
| 1146 |
int type = !memcmp(map->volume, MAPPER_PREFIX, MAPPER_PREFIX_LEN);
|
| 1147 |
XSEGLOG2(&lc, I, "Type %d detected for map %s", type, map->volume);
|
| 1148 |
uint32_t version; |
| 1149 |
if (type)
|
| 1150 |
version = *(uint32_t *) (buf); //version should always be the first uint32_t
|
| 1151 |
else
|
| 1152 |
version = 0;
|
| 1153 |
if (version > MAP_LATEST_VERSION){
|
| 1154 |
XSEGLOG2(&lc, E, "Map read for map %s failed. Invalid version %u",
|
| 1155 |
map->volume, version); |
| 1156 |
return -1; |
| 1157 |
} |
| 1158 |
|
| 1159 |
r = map_functions[version].read_map(map, buf); |
| 1160 |
if (r < 0){ |
| 1161 |
XSEGLOG2(&lc, E, "Map read for map %s failed", map->volume);
|
| 1162 |
return -1; |
| 1163 |
} |
| 1164 |
|
| 1165 |
print_map(map); |
| 1166 |
XSEGLOG2(&lc, I, "Map read for map %s completed", map->volume);
|
| 1167 |
return 0; |
| 1168 |
|
| 1169 |
} |
| 1170 |
|
| 1171 |
static int load_map(struct peer_req *pr, struct map *map) |
| 1172 |
{
|
| 1173 |
int r = 0; |
| 1174 |
struct xseg_request *req;
|
| 1175 |
struct peerd *peer = pr->peer;
|
| 1176 |
req = __load_map(pr, map); |
| 1177 |
if (!req)
|
| 1178 |
return -1; |
| 1179 |
wait_on_pr(pr, (!(req->state & XS_FAILED || req->state & XS_SERVED))); |
| 1180 |
map->flags &= ~MF_MAP_LOADING; |
| 1181 |
if (req->state & XS_FAILED){
|
| 1182 |
XSEGLOG2(&lc, E, "Map load failed for map %s", map->volume);
|
| 1183 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1184 |
return -1; |
| 1185 |
} |
| 1186 |
r = read_map(map, (unsigned char *) xseg_get_data(peer->xseg, req)); |
| 1187 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1188 |
return r;
|
| 1189 |
} |
| 1190 |
|
| 1191 |
static struct xseg_request * __open_map(struct peer_req *pr, struct map *m, |
| 1192 |
uint32_t flags) |
| 1193 |
{
|
| 1194 |
int r;
|
| 1195 |
xport p; |
| 1196 |
struct xseg_request *req;
|
| 1197 |
struct peerd *peer = pr->peer;
|
| 1198 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1199 |
void *dummy;
|
| 1200 |
|
| 1201 |
XSEGLOG2(&lc, I, "Opening map %s", m->volume);
|
| 1202 |
|
| 1203 |
req = xseg_get_request(peer->xseg, pr->portno, mapper->mbportno, X_ALLOC); |
| 1204 |
if (!req){
|
| 1205 |
XSEGLOG2(&lc, E, "Cannot allocate request for map %s",
|
| 1206 |
m->volume); |
| 1207 |
goto out_fail;
|
| 1208 |
} |
| 1209 |
|
| 1210 |
r = xseg_prep_request(peer->xseg, req, m->volumelen, block_size); |
| 1211 |
if (r < 0){ |
| 1212 |
XSEGLOG2(&lc, E, "Cannot prepare request for map %s",
|
| 1213 |
m->volume); |
| 1214 |
goto out_put;
|
| 1215 |
} |
| 1216 |
|
| 1217 |
char *reqtarget = xseg_get_target(peer->xseg, req);
|
| 1218 |
if (!reqtarget)
|
| 1219 |
goto out_put;
|
| 1220 |
strncpy(reqtarget, m->volume, req->targetlen); |
| 1221 |
req->op = X_ACQUIRE; |
| 1222 |
req->size = block_size; |
| 1223 |
req->offset = 0;
|
| 1224 |
if (!(flags & MF_FORCE))
|
| 1225 |
req->flags = XF_NOSYNC; |
| 1226 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 1227 |
if (r < 0){ |
| 1228 |
XSEGLOG2(&lc, E, "Cannot set request data for map %s",
|
| 1229 |
m->volume); |
| 1230 |
goto out_put;
|
| 1231 |
} |
| 1232 |
p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 1233 |
if (p == NoPort){
|
| 1234 |
XSEGLOG2(&lc, E, "Cannot submit request for map %s",
|
| 1235 |
m->volume); |
| 1236 |
goto out_unset;
|
| 1237 |
} |
| 1238 |
r = xseg_signal(peer->xseg, p); |
| 1239 |
|
| 1240 |
m->flags |= MF_MAP_OPENING; |
| 1241 |
XSEGLOG2(&lc, I, "Map %s opening", m->volume);
|
| 1242 |
return req;
|
| 1243 |
|
| 1244 |
out_unset:
|
| 1245 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 1246 |
out_put:
|
| 1247 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1248 |
out_fail:
|
| 1249 |
return NULL; |
| 1250 |
} |
| 1251 |
|
| 1252 |
static int open_map(struct peer_req *pr, struct map *map, uint32_t flags) |
| 1253 |
{
|
| 1254 |
int err;
|
| 1255 |
struct xseg_request *req;
|
| 1256 |
struct peerd *peer = pr->peer;
|
| 1257 |
|
| 1258 |
req = __open_map(pr, map, flags); |
| 1259 |
if (!req){
|
| 1260 |
return -1; |
| 1261 |
} |
| 1262 |
wait_on_pr(pr, (!((req->state & XS_FAILED)||(req->state & XS_SERVED)))); |
| 1263 |
map->flags &= ~MF_MAP_OPENING; |
| 1264 |
err = req->state & XS_FAILED; |
| 1265 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1266 |
if (err)
|
| 1267 |
return -1; |
| 1268 |
else
|
| 1269 |
map->flags |= MF_MAP_EXCLUSIVE; |
| 1270 |
return 0; |
| 1271 |
} |
| 1272 |
|
| 1273 |
/*
|
| 1274 |
* copy up functions
|
| 1275 |
*/
|
| 1276 |
|
| 1277 |
static int __set_copyup_node(struct mapper_io *mio, struct xseg_request *req, struct map_node *mn) |
| 1278 |
{
|
| 1279 |
int r = 0; |
| 1280 |
if (mn){
|
| 1281 |
XSEGLOG2(&lc, D, "Inserting (req: %lx, mapnode: %lx) on mio %lx",
|
| 1282 |
req, mn, mio); |
| 1283 |
r = xhash_insert(mio->copyups_nodes, (xhashidx) req, (xhashidx) mn); |
| 1284 |
if (r == -XHASH_ERESIZE) {
|
| 1285 |
xhashidx shift = xhash_grow_size_shift(mio->copyups_nodes); |
| 1286 |
xhash_t *new_hashmap = xhash_resize(mio->copyups_nodes, shift, 0, NULL); |
| 1287 |
if (!new_hashmap)
|
| 1288 |
goto out;
|
| 1289 |
mio->copyups_nodes = new_hashmap; |
| 1290 |
r = xhash_insert(mio->copyups_nodes, (xhashidx) req, (xhashidx) mn); |
| 1291 |
} |
| 1292 |
if (r < 0) |
| 1293 |
XSEGLOG2(&lc, E, "Insertion of (%lx, %lx) on mio %lx failed",
|
| 1294 |
req, mn, mio); |
| 1295 |
} |
| 1296 |
else {
|
| 1297 |
XSEGLOG2(&lc, D, "Deleting req: %lx from mio %lx",
|
| 1298 |
req, mio); |
| 1299 |
r = xhash_delete(mio->copyups_nodes, (xhashidx) req); |
| 1300 |
if (r == -XHASH_ERESIZE) {
|
| 1301 |
xhashidx shift = xhash_shrink_size_shift(mio->copyups_nodes); |
| 1302 |
xhash_t *new_hashmap = xhash_resize(mio->copyups_nodes, shift, 0, NULL); |
| 1303 |
if (!new_hashmap)
|
| 1304 |
goto out;
|
| 1305 |
mio->copyups_nodes = new_hashmap; |
| 1306 |
r = xhash_delete(mio->copyups_nodes, (xhashidx) req); |
| 1307 |
} |
| 1308 |
if (r < 0) |
| 1309 |
XSEGLOG2(&lc, E, "Deletion of %lx on mio %lx failed",
|
| 1310 |
req, mio); |
| 1311 |
} |
| 1312 |
out:
|
| 1313 |
return r;
|
| 1314 |
} |
| 1315 |
|
| 1316 |
static struct map_node * __get_copyup_node(struct mapper_io *mio, struct xseg_request *req) |
| 1317 |
{
|
| 1318 |
struct map_node *mn;
|
| 1319 |
int r = xhash_lookup(mio->copyups_nodes, (xhashidx) req, (xhashidx *) &mn);
|
| 1320 |
if (r < 0){ |
| 1321 |
XSEGLOG2(&lc, W, "Cannot find req %lx on mio %lx", req, mio);
|
| 1322 |
return NULL; |
| 1323 |
} |
| 1324 |
XSEGLOG2(&lc, D, "Found mapnode %lx req %lx on mio %lx", mn, req, mio);
|
| 1325 |
return mn;
|
| 1326 |
} |
| 1327 |
|
| 1328 |
static struct xseg_request * __snapshot_object(struct peer_req *pr, |
| 1329 |
struct map_node *mn)
|
| 1330 |
{
|
| 1331 |
struct peerd *peer = pr->peer;
|
| 1332 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1333 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 1334 |
//struct map *map = mn->map;
|
| 1335 |
void *dummy;
|
| 1336 |
int r = -1; |
| 1337 |
xport p; |
| 1338 |
|
| 1339 |
//assert mn->volume != zero_block
|
| 1340 |
//assert mn->flags & MF_OBJECT_EXIST
|
| 1341 |
struct xseg_request *req = xseg_get_request(peer->xseg, pr->portno,
|
| 1342 |
mapper->bportno, X_ALLOC); |
| 1343 |
if (!req){
|
| 1344 |
XSEGLOG2(&lc, E, "Cannot get request for object %s", mn->object);
|
| 1345 |
goto out_err;
|
| 1346 |
} |
| 1347 |
r = xseg_prep_request(peer->xseg, req, mn->objectlen, |
| 1348 |
sizeof(struct xseg_request_snapshot)); |
| 1349 |
if (r < 0){ |
| 1350 |
XSEGLOG2(&lc, E, "Cannot prepare request for object %s", mn->object);
|
| 1351 |
goto out_put;
|
| 1352 |
} |
| 1353 |
|
| 1354 |
char *target = xseg_get_target(peer->xseg, req);
|
| 1355 |
strncpy(target, mn->object, req->targetlen); |
| 1356 |
|
| 1357 |
struct xseg_request_snapshot *xsnapshot = (struct xseg_request_snapshot *) xseg_get_data(peer->xseg, req); |
| 1358 |
xsnapshot->target[0] = 0; |
| 1359 |
xsnapshot->targetlen = 0;
|
| 1360 |
|
| 1361 |
req->offset = 0;
|
| 1362 |
req->size = block_size; |
| 1363 |
req->op = X_SNAPSHOT; |
| 1364 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 1365 |
if (r<0){ |
| 1366 |
XSEGLOG2(&lc, E, "Cannot set request data for object %s", mn->object);
|
| 1367 |
goto out_put;
|
| 1368 |
} |
| 1369 |
r = __set_copyup_node(mio, req, mn); |
| 1370 |
if (r < 0) |
| 1371 |
goto out_unset;
|
| 1372 |
p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 1373 |
if (p == NoPort) {
|
| 1374 |
XSEGLOG2(&lc, E, "Cannot submit for object %s", mn->object);
|
| 1375 |
goto out_mapper_unset;
|
| 1376 |
} |
| 1377 |
xseg_signal(peer->xseg, p); |
| 1378 |
|
| 1379 |
mn->flags |= MF_OBJECT_SNAPSHOTTING; |
| 1380 |
XSEGLOG2(&lc, I, "Snapshotting up object %s", mn->object);
|
| 1381 |
return req;
|
| 1382 |
|
| 1383 |
out_mapper_unset:
|
| 1384 |
__set_copyup_node(mio, req, NULL);
|
| 1385 |
out_unset:
|
| 1386 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 1387 |
out_put:
|
| 1388 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1389 |
out_err:
|
| 1390 |
XSEGLOG2(&lc, E, "Snapshotting object %s failed", mn->object);
|
| 1391 |
return NULL; |
| 1392 |
} |
| 1393 |
|
| 1394 |
static struct xseg_request * copyup_object(struct peerd *peer, struct map_node *mn, struct peer_req *pr) |
| 1395 |
{
|
| 1396 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1397 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 1398 |
struct map *map = mn->map;
|
| 1399 |
void *dummy;
|
| 1400 |
int r = -1; |
| 1401 |
xport p; |
| 1402 |
|
| 1403 |
uint32_t newtargetlen; |
| 1404 |
char new_target[MAX_OBJECT_LEN + 1]; |
| 1405 |
unsigned char sha[SHA256_DIGEST_SIZE]; |
| 1406 |
|
| 1407 |
strncpy(new_target, MAPPER_PREFIX, MAPPER_PREFIX_LEN); |
| 1408 |
|
| 1409 |
char tmp[XSEG_MAX_TARGETLEN + 1]; |
| 1410 |
uint32_t tmplen; |
| 1411 |
strncpy(tmp, map->volume, map->volumelen); |
| 1412 |
sprintf(tmp + map->volumelen, "_%u", mn->objectidx);
|
| 1413 |
tmp[XSEG_MAX_TARGETLEN] = 0;
|
| 1414 |
tmplen = strlen(tmp); |
| 1415 |
SHA256((unsigned char *)tmp, tmplen, sha); |
| 1416 |
hexlify(sha, new_target+MAPPER_PREFIX_LEN); |
| 1417 |
newtargetlen = MAPPER_PREFIX_LEN + HEXLIFIED_SHA256_DIGEST_SIZE; |
| 1418 |
|
| 1419 |
|
| 1420 |
if (!strncmp(mn->object, zero_block, ZERO_BLOCK_LEN))
|
| 1421 |
goto copyup_zeroblock;
|
| 1422 |
|
| 1423 |
struct xseg_request *req = xseg_get_request(peer->xseg, pr->portno,
|
| 1424 |
mapper->bportno, X_ALLOC); |
| 1425 |
if (!req){
|
| 1426 |
XSEGLOG2(&lc, E, "Cannot get request for object %s", mn->object);
|
| 1427 |
goto out_err;
|
| 1428 |
} |
| 1429 |
r = xseg_prep_request(peer->xseg, req, newtargetlen, |
| 1430 |
sizeof(struct xseg_request_copy)); |
| 1431 |
if (r < 0){ |
| 1432 |
XSEGLOG2(&lc, E, "Cannot prepare request for object %s", mn->object);
|
| 1433 |
goto out_put;
|
| 1434 |
} |
| 1435 |
|
| 1436 |
char *target = xseg_get_target(peer->xseg, req);
|
| 1437 |
strncpy(target, new_target, req->targetlen); |
| 1438 |
|
| 1439 |
struct xseg_request_copy *xcopy = (struct xseg_request_copy *) xseg_get_data(peer->xseg, req); |
| 1440 |
strncpy(xcopy->target, mn->object, mn->objectlen); |
| 1441 |
xcopy->targetlen = mn->objectlen; |
| 1442 |
|
| 1443 |
req->offset = 0;
|
| 1444 |
req->size = block_size; |
| 1445 |
req->op = X_COPY; |
| 1446 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 1447 |
if (r<0){ |
| 1448 |
XSEGLOG2(&lc, E, "Cannot set request data for object %s", mn->object);
|
| 1449 |
goto out_put;
|
| 1450 |
} |
| 1451 |
r = __set_copyup_node(mio, req, mn); |
| 1452 |
if (r < 0) |
| 1453 |
goto out_unset;
|
| 1454 |
p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 1455 |
if (p == NoPort) {
|
| 1456 |
XSEGLOG2(&lc, E, "Cannot submit for object %s", mn->object);
|
| 1457 |
goto out_mapper_unset;
|
| 1458 |
} |
| 1459 |
xseg_signal(peer->xseg, p); |
| 1460 |
// mio->copyups++;
|
| 1461 |
|
| 1462 |
mn->flags |= MF_OBJECT_COPYING; |
| 1463 |
XSEGLOG2(&lc, I, "Copying up object %s \n\t to %s", mn->object, new_target);
|
| 1464 |
return req;
|
| 1465 |
|
| 1466 |
out_mapper_unset:
|
| 1467 |
__set_copyup_node(mio, req, NULL);
|
| 1468 |
out_unset:
|
| 1469 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 1470 |
out_put:
|
| 1471 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1472 |
out_err:
|
| 1473 |
XSEGLOG2(&lc, E, "Copying up object %s \n\t to %s failed", mn->object, new_target);
|
| 1474 |
return NULL; |
| 1475 |
|
| 1476 |
copyup_zeroblock:
|
| 1477 |
XSEGLOG2(&lc, I, "Copying up of zero block is not needed."
|
| 1478 |
"Proceeding in writing the new object in map");
|
| 1479 |
/* construct a tmp map_node for writing purposes */
|
| 1480 |
struct map_node newmn = *mn;
|
| 1481 |
newmn.flags = MF_OBJECT_EXIST; |
| 1482 |
strncpy(newmn.object, new_target, newtargetlen); |
| 1483 |
newmn.object[newtargetlen] = 0;
|
| 1484 |
newmn.objectlen = newtargetlen; |
| 1485 |
newmn.objectidx = mn->objectidx; |
| 1486 |
req = object_write(peer, pr, map, &newmn); |
| 1487 |
r = __set_copyup_node(mio, req, mn); |
| 1488 |
if (r < 0) |
| 1489 |
return NULL; |
| 1490 |
if (!req){
|
| 1491 |
XSEGLOG2(&lc, E, "Object write returned error for object %s"
|
| 1492 |
"\n\t of map %s [%llu]",
|
| 1493 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 1494 |
return NULL; |
| 1495 |
} |
| 1496 |
mn->flags |= MF_OBJECT_WRITING; |
| 1497 |
XSEGLOG2(&lc, I, "Object %s copy up completed. Pending writing.", mn->object);
|
| 1498 |
return req;
|
| 1499 |
} |
| 1500 |
|
| 1501 |
static struct xseg_request * __delete_object(struct peer_req *pr, struct map_node *mn) |
| 1502 |
{
|
| 1503 |
void *dummy;
|
| 1504 |
struct peerd *peer = pr->peer;
|
| 1505 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1506 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 1507 |
struct xseg_request *req = xseg_get_request(peer->xseg, pr->portno,
|
| 1508 |
mapper->bportno, X_ALLOC); |
| 1509 |
XSEGLOG2(&lc, I, "Deleting mapnode %s", mn->object);
|
| 1510 |
if (!req){
|
| 1511 |
XSEGLOG2(&lc, E, "Cannot get request for object %s", mn->object);
|
| 1512 |
goto out_err;
|
| 1513 |
} |
| 1514 |
int r = xseg_prep_request(peer->xseg, req, mn->objectlen, 0); |
| 1515 |
if (r < 0){ |
| 1516 |
XSEGLOG2(&lc, E, "Cannot prep request for object %s", mn->object);
|
| 1517 |
goto out_put;
|
| 1518 |
} |
| 1519 |
char *target = xseg_get_target(peer->xseg, req);
|
| 1520 |
strncpy(target, mn->object, req->targetlen); |
| 1521 |
req->op = X_DELETE; |
| 1522 |
req->size = req->datalen; |
| 1523 |
req->offset = 0;
|
| 1524 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 1525 |
if (r < 0){ |
| 1526 |
XSEGLOG2(&lc, E, "Cannot set req data for object %s", mn->object);
|
| 1527 |
goto out_put;
|
| 1528 |
} |
| 1529 |
r = __set_copyup_node(mio, req, mn); |
| 1530 |
if (r < 0) |
| 1531 |
goto out_unset;
|
| 1532 |
xport p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 1533 |
if (p == NoPort){
|
| 1534 |
XSEGLOG2(&lc, E, "Cannot submit request for object %s", mn->object);
|
| 1535 |
goto out_mapper_unset;
|
| 1536 |
} |
| 1537 |
r = xseg_signal(peer->xseg, p); |
| 1538 |
mn->flags |= MF_OBJECT_DELETING; |
| 1539 |
XSEGLOG2(&lc, I, "Object %s deletion pending", mn->object);
|
| 1540 |
return req;
|
| 1541 |
|
| 1542 |
out_mapper_unset:
|
| 1543 |
__set_copyup_node(mio, req, NULL);
|
| 1544 |
out_unset:
|
| 1545 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 1546 |
out_put:
|
| 1547 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1548 |
out_err:
|
| 1549 |
XSEGLOG2(&lc, I, "Object %s deletion failed", mn->object);
|
| 1550 |
return NULL; |
| 1551 |
} |
| 1552 |
|
| 1553 |
static struct xseg_request * __delete_map(struct peer_req *pr, struct map *map) |
| 1554 |
{
|
| 1555 |
void *dummy;
|
| 1556 |
struct peerd *peer = pr->peer;
|
| 1557 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1558 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 1559 |
struct xseg_request *req = xseg_get_request(peer->xseg, pr->portno,
|
| 1560 |
mapper->mbportno, X_ALLOC); |
| 1561 |
XSEGLOG2(&lc, I, "Deleting map %s", map->volume);
|
| 1562 |
if (!req){
|
| 1563 |
XSEGLOG2(&lc, E, "Cannot get request for map %s", map->volume);
|
| 1564 |
goto out_err;
|
| 1565 |
} |
| 1566 |
int r = xseg_prep_request(peer->xseg, req, map->volumelen, 0); |
| 1567 |
if (r < 0){ |
| 1568 |
XSEGLOG2(&lc, E, "Cannot prep request for map %s", map->volume);
|
| 1569 |
goto out_put;
|
| 1570 |
} |
| 1571 |
char *target = xseg_get_target(peer->xseg, req);
|
| 1572 |
strncpy(target, map->volume, req->targetlen); |
| 1573 |
req->op = X_DELETE; |
| 1574 |
req->size = req->datalen; |
| 1575 |
req->offset = 0;
|
| 1576 |
r = xseg_set_req_data(peer->xseg, req, pr); |
| 1577 |
if (r < 0){ |
| 1578 |
XSEGLOG2(&lc, E, "Cannot set req data for map %s", map->volume);
|
| 1579 |
goto out_put;
|
| 1580 |
} |
| 1581 |
/* do not check return value. just make sure there is no node set */
|
| 1582 |
__set_copyup_node(mio, req, NULL);
|
| 1583 |
xport p = xseg_submit(peer->xseg, req, pr->portno, X_ALLOC); |
| 1584 |
if (p == NoPort){
|
| 1585 |
XSEGLOG2(&lc, E, "Cannot submit request for map %s", map->volume);
|
| 1586 |
goto out_unset;
|
| 1587 |
} |
| 1588 |
r = xseg_signal(peer->xseg, p); |
| 1589 |
map->flags |= MF_MAP_DELETING; |
| 1590 |
XSEGLOG2(&lc, I, "Map %s deletion pending", map->volume);
|
| 1591 |
return req;
|
| 1592 |
|
| 1593 |
out_unset:
|
| 1594 |
xseg_get_req_data(peer->xseg, req, &dummy); |
| 1595 |
out_put:
|
| 1596 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1597 |
out_err:
|
| 1598 |
XSEGLOG2(&lc, E, "Map %s deletion failed", map->volume);
|
| 1599 |
return NULL; |
| 1600 |
} |
| 1601 |
|
| 1602 |
|
| 1603 |
static inline struct map_node * get_mapnode(struct map *map, uint32_t index) |
| 1604 |
{
|
| 1605 |
struct map_node *mn = find_object(map, index);
|
| 1606 |
if (mn)
|
| 1607 |
mn->ref++; |
| 1608 |
return mn;
|
| 1609 |
} |
| 1610 |
|
| 1611 |
static inline void put_mapnode(struct map_node *mn) |
| 1612 |
{
|
| 1613 |
mn->ref--; |
| 1614 |
if (!mn->ref){
|
| 1615 |
//clean up mn
|
| 1616 |
st_cond_destroy(mn->cond); |
| 1617 |
} |
| 1618 |
} |
| 1619 |
|
| 1620 |
static inline void __get_map(struct map *map) |
| 1621 |
{
|
| 1622 |
map->ref++; |
| 1623 |
} |
| 1624 |
|
| 1625 |
static inline void put_map(struct map *map) |
| 1626 |
{
|
| 1627 |
struct map_node *mn;
|
| 1628 |
map->ref--; |
| 1629 |
if (!map->ref){
|
| 1630 |
XSEGLOG2(&lc, I, "Freeing map %s", map->volume);
|
| 1631 |
//clean up map
|
| 1632 |
uint64_t i; |
| 1633 |
for (i = 0; i < calc_map_obj(map); i++) { |
| 1634 |
mn = get_mapnode(map, i); |
| 1635 |
if (mn) {
|
| 1636 |
//make sure all pending operations on all objects are completed
|
| 1637 |
//this should never happen...
|
| 1638 |
wait_on_mapnode(mn, mn->flags & MF_OBJECT_NOT_READY); |
| 1639 |
mn->flags |= MF_OBJECT_DESTROYED; |
| 1640 |
put_mapnode(mn); //matchin mn->ref = 1 on mn init
|
| 1641 |
put_mapnode(mn); //matcing get_mapnode;
|
| 1642 |
//assert mn->ref == 0;
|
| 1643 |
} |
| 1644 |
} |
| 1645 |
mn = find_object(map, 0);
|
| 1646 |
if (mn)
|
| 1647 |
free(mn); |
| 1648 |
XSEGLOG2(&lc, I, "Freed map %s", map->volume);
|
| 1649 |
free(map); |
| 1650 |
} |
| 1651 |
} |
| 1652 |
|
| 1653 |
static struct map * create_map(struct mapperd *mapper, char *name, |
| 1654 |
uint32_t namelen, uint32_t flags) |
| 1655 |
{
|
| 1656 |
int r;
|
| 1657 |
if (namelen + MAPPER_PREFIX_LEN > MAX_VOLUME_LEN){
|
| 1658 |
XSEGLOG2(&lc, E, "Namelen %u too long. Max: %d",
|
| 1659 |
namelen, MAX_VOLUME_LEN); |
| 1660 |
return NULL; |
| 1661 |
} |
| 1662 |
struct map *m = malloc(sizeof(struct map)); |
| 1663 |
if (!m){
|
| 1664 |
XSEGLOG2(&lc, E, "Cannot allocate map ");
|
| 1665 |
goto out_err;
|
| 1666 |
} |
| 1667 |
m->size = -1;
|
| 1668 |
if (flags & MF_ARCHIP){
|
| 1669 |
strncpy(m->volume, MAPPER_PREFIX, MAPPER_PREFIX_LEN); |
| 1670 |
strncpy(m->volume + MAPPER_PREFIX_LEN, name, namelen); |
| 1671 |
m->volume[MAPPER_PREFIX_LEN + namelen] = 0;
|
| 1672 |
m->volumelen = MAPPER_PREFIX_LEN + namelen; |
| 1673 |
m->version = 1; /* keep this hardcoded for now */ |
| 1674 |
} |
| 1675 |
else {
|
| 1676 |
strncpy(m->volume, name, namelen); |
| 1677 |
m->volume[namelen] = 0;
|
| 1678 |
m->volumelen = namelen; |
| 1679 |
m->version = 0; /* version 0 should be pithos maps */ |
| 1680 |
} |
| 1681 |
m->flags = 0;
|
| 1682 |
m->objects = xhash_new(3, 0, INTEGER); |
| 1683 |
if (!m->objects){
|
| 1684 |
XSEGLOG2(&lc, E, "Cannot allocate object hashmap for map %s",
|
| 1685 |
m->volume); |
| 1686 |
goto out_map;
|
| 1687 |
} |
| 1688 |
m->ref = 1;
|
| 1689 |
m->waiters = 0;
|
| 1690 |
m->cond = st_cond_new(); //FIXME err check;
|
| 1691 |
r = insert_map(mapper, m); |
| 1692 |
if (r < 0){ |
| 1693 |
XSEGLOG2(&lc, E, "Cannot insert map %s", m->volume);
|
| 1694 |
goto out_hash;
|
| 1695 |
} |
| 1696 |
|
| 1697 |
return m;
|
| 1698 |
|
| 1699 |
out_hash:
|
| 1700 |
xhash_free(m->objects); |
| 1701 |
out_map:
|
| 1702 |
XSEGLOG2(&lc, E, "failed to create map %s", m->volume);
|
| 1703 |
free(m); |
| 1704 |
out_err:
|
| 1705 |
return NULL; |
| 1706 |
} |
| 1707 |
|
| 1708 |
|
| 1709 |
|
| 1710 |
void deletion_cb(struct peer_req *pr, struct xseg_request *req) |
| 1711 |
{
|
| 1712 |
struct peerd *peer = pr->peer;
|
| 1713 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1714 |
(void)mapper;
|
| 1715 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 1716 |
struct map_node *mn = __get_copyup_node(mio, req);
|
| 1717 |
|
| 1718 |
__set_copyup_node(mio, req, NULL);
|
| 1719 |
|
| 1720 |
//assert req->op = X_DELETE;
|
| 1721 |
//assert pr->req->op = X_DELETE only map deletions make delete requests
|
| 1722 |
//assert mio->del_pending > 0
|
| 1723 |
XSEGLOG2(&lc, D, "mio: %lx, del_pending: %llu", mio, mio->del_pending);
|
| 1724 |
mio->del_pending--; |
| 1725 |
|
| 1726 |
if (req->state & XS_FAILED){
|
| 1727 |
mio->err = 1;
|
| 1728 |
} |
| 1729 |
if (mn){
|
| 1730 |
XSEGLOG2(&lc, D, "Found mapnode %lx %s for mio: %lx, req: %lx",
|
| 1731 |
mn, mn->object, mio, req); |
| 1732 |
// assert mn->flags & MF_OBJECT_DELETING
|
| 1733 |
mn->flags &= ~MF_OBJECT_DELETING; |
| 1734 |
mn->flags |= MF_OBJECT_DESTROYED; |
| 1735 |
signal_mapnode(mn); |
| 1736 |
/* put mapnode here, matches get_mapnode on do_destroy */
|
| 1737 |
put_mapnode(mn); |
| 1738 |
} else {
|
| 1739 |
XSEGLOG2(&lc, E, "Cannot get map node for mio: %lx, req: %lx",
|
| 1740 |
mio, req); |
| 1741 |
} |
| 1742 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1743 |
signal_pr(pr); |
| 1744 |
} |
| 1745 |
|
| 1746 |
void snapshot_cb(struct peer_req *pr, struct xseg_request *req) |
| 1747 |
{
|
| 1748 |
struct peerd *peer = pr->peer;
|
| 1749 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1750 |
(void)mapper;
|
| 1751 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 1752 |
struct map_node *mn = __get_copyup_node(mio, req);
|
| 1753 |
if (!mn){
|
| 1754 |
XSEGLOG2(&lc, E, "Cannot get map node");
|
| 1755 |
goto out_err;
|
| 1756 |
} |
| 1757 |
__set_copyup_node(mio, req, NULL);
|
| 1758 |
|
| 1759 |
if (req->state & XS_FAILED){
|
| 1760 |
if (req->op == X_DELETE){
|
| 1761 |
XSEGLOG2(&lc, E, "Delete req failed");
|
| 1762 |
goto out_ok;
|
| 1763 |
} |
| 1764 |
XSEGLOG2(&lc, E, "Req failed");
|
| 1765 |
mn->flags &= ~MF_OBJECT_SNAPSHOTTING; |
| 1766 |
mn->flags &= ~MF_OBJECT_WRITING; |
| 1767 |
goto out_err;
|
| 1768 |
} |
| 1769 |
|
| 1770 |
if (req->op == X_WRITE) {
|
| 1771 |
char old_object_name[MAX_OBJECT_LEN + 1]; |
| 1772 |
uint32_t old_objectlen; |
| 1773 |
|
| 1774 |
char *target = xseg_get_target(peer->xseg, req);
|
| 1775 |
(void)target;
|
| 1776 |
//assert mn->flags & MF_OBJECT_WRITING
|
| 1777 |
mn->flags &= ~MF_OBJECT_WRITING; |
| 1778 |
strncpy(old_object_name, mn->object, mn->objectlen); |
| 1779 |
old_objectlen = mn->objectlen; |
| 1780 |
|
| 1781 |
struct map_node tmp;
|
| 1782 |
char *data = xseg_get_data(peer->xseg, req);
|
| 1783 |
map_to_object(&tmp, (unsigned char *) data); |
| 1784 |
mn->flags &= ~MF_OBJECT_EXIST; |
| 1785 |
|
| 1786 |
strncpy(mn->object, tmp.object, tmp.objectlen); |
| 1787 |
mn->object[tmp.objectlen] = 0;
|
| 1788 |
mn->objectlen = tmp.objectlen; |
| 1789 |
XSEGLOG2(&lc, I, "Object write of %s completed successfully", mn->object);
|
| 1790 |
//signal_mapnode since Snapshot was successfull
|
| 1791 |
signal_mapnode(mn); |
| 1792 |
|
| 1793 |
//do delete old object
|
| 1794 |
strncpy(tmp.object, old_object_name, old_objectlen); |
| 1795 |
tmp.object[old_objectlen] = 0;
|
| 1796 |
tmp.objectlen = old_objectlen; |
| 1797 |
tmp.flags = MF_OBJECT_EXIST; |
| 1798 |
struct xseg_request *xreq = __delete_object(pr, &tmp);
|
| 1799 |
if (!xreq){
|
| 1800 |
//just a warning. Snapshot was successfull
|
| 1801 |
XSEGLOG2(&lc, W, "Cannot delete old object %s", tmp.object);
|
| 1802 |
goto out_ok;
|
| 1803 |
} |
| 1804 |
//overwrite copyup node, since tmp is a stack dummy variable
|
| 1805 |
__set_copyup_node (mio, xreq, mn); |
| 1806 |
XSEGLOG2(&lc, I, "Deletion of %s pending", tmp.object);
|
| 1807 |
} else if (req->op == X_SNAPSHOT) { |
| 1808 |
//issue write_object;
|
| 1809 |
mn->flags &= ~MF_OBJECT_SNAPSHOTTING; |
| 1810 |
struct map *map = mn->map;
|
| 1811 |
if (!map){
|
| 1812 |
XSEGLOG2(&lc, E, "Object %s has not map back pointer", mn->object);
|
| 1813 |
goto out_err;
|
| 1814 |
} |
| 1815 |
|
| 1816 |
/* construct a tmp map_node for writing purposes */
|
| 1817 |
//char *target = xseg_get_target(peer->xseg, req);
|
| 1818 |
struct map_node newmn = *mn;
|
| 1819 |
newmn.flags = 0;
|
| 1820 |
struct xseg_reply_snapshot *xreply;
|
| 1821 |
xreply = (struct xseg_reply_snapshot *) xseg_get_data(peer->xseg, req);
|
| 1822 |
//assert xreply->targetlen !=0
|
| 1823 |
//assert xreply->targetlen < XSEG_MAX_TARGETLEN
|
| 1824 |
//xreply->target[xreply->targetlen] = 0;
|
| 1825 |
//assert xreply->target valid
|
| 1826 |
strncpy(newmn.object, xreply->target, xreply->targetlen); |
| 1827 |
newmn.object[req->targetlen] = 0;
|
| 1828 |
newmn.objectlen = req->targetlen; |
| 1829 |
newmn.objectidx = mn->objectidx; |
| 1830 |
struct xseg_request *xreq = object_write(peer, pr, map, &newmn);
|
| 1831 |
if (!xreq){
|
| 1832 |
XSEGLOG2(&lc, E, "Object write returned error for object %s"
|
| 1833 |
"\n\t of map %s [%llu]",
|
| 1834 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 1835 |
goto out_err;
|
| 1836 |
} |
| 1837 |
mn->flags |= MF_OBJECT_WRITING; |
| 1838 |
__set_copyup_node (mio, xreq, mn); |
| 1839 |
|
| 1840 |
XSEGLOG2(&lc, I, "Object %s snapshot completed. Pending writing.", mn->object);
|
| 1841 |
} else if (req->op == X_DELETE){ |
| 1842 |
//deletion of the old block completed
|
| 1843 |
XSEGLOG2(&lc, I, "Deletion of completed");
|
| 1844 |
goto out_ok;
|
| 1845 |
; |
| 1846 |
} else {
|
| 1847 |
//wtf??
|
| 1848 |
; |
| 1849 |
} |
| 1850 |
|
| 1851 |
out:
|
| 1852 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1853 |
return;
|
| 1854 |
|
| 1855 |
out_err:
|
| 1856 |
mio->snap_pending--; |
| 1857 |
XSEGLOG2(&lc, D, "Mio->snap_pending: %u", mio->snap_pending);
|
| 1858 |
mio->err = 1;
|
| 1859 |
if (mn)
|
| 1860 |
signal_mapnode(mn); |
| 1861 |
signal_pr(pr); |
| 1862 |
goto out;
|
| 1863 |
|
| 1864 |
out_ok:
|
| 1865 |
mio->snap_pending--; |
| 1866 |
signal_pr(pr); |
| 1867 |
goto out;
|
| 1868 |
|
| 1869 |
|
| 1870 |
} |
| 1871 |
void copyup_cb(struct peer_req *pr, struct xseg_request *req) |
| 1872 |
{
|
| 1873 |
struct peerd *peer = pr->peer;
|
| 1874 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 1875 |
(void)mapper;
|
| 1876 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 1877 |
struct map_node *mn = __get_copyup_node(mio, req);
|
| 1878 |
if (!mn){
|
| 1879 |
XSEGLOG2(&lc, E, "Cannot get map node");
|
| 1880 |
goto out_err;
|
| 1881 |
} |
| 1882 |
__set_copyup_node(mio, req, NULL);
|
| 1883 |
|
| 1884 |
if (req->state & XS_FAILED){
|
| 1885 |
XSEGLOG2(&lc, E, "Req failed");
|
| 1886 |
mn->flags &= ~MF_OBJECT_COPYING; |
| 1887 |
mn->flags &= ~MF_OBJECT_WRITING; |
| 1888 |
goto out_err;
|
| 1889 |
} |
| 1890 |
if (req->op == X_WRITE) {
|
| 1891 |
char *target = xseg_get_target(peer->xseg, req);
|
| 1892 |
(void)target;
|
| 1893 |
//printf("handle object write replyi\n");
|
| 1894 |
__set_copyup_node(mio, req, NULL);
|
| 1895 |
//assert mn->flags & MF_OBJECT_WRITING
|
| 1896 |
mn->flags &= ~MF_OBJECT_WRITING; |
| 1897 |
|
| 1898 |
struct map_node tmp;
|
| 1899 |
char *data = xseg_get_data(peer->xseg, req);
|
| 1900 |
map_to_object(&tmp, (unsigned char *) data); |
| 1901 |
mn->flags |= MF_OBJECT_EXIST; |
| 1902 |
if (mn->flags != MF_OBJECT_EXIST){
|
| 1903 |
XSEGLOG2(&lc, E, "map node %s has wrong flags", mn->object);
|
| 1904 |
goto out_err;
|
| 1905 |
} |
| 1906 |
//assert mn->flags & MF_OBJECT_EXIST
|
| 1907 |
strncpy(mn->object, tmp.object, tmp.objectlen); |
| 1908 |
mn->object[tmp.objectlen] = 0;
|
| 1909 |
mn->objectlen = tmp.objectlen; |
| 1910 |
XSEGLOG2(&lc, I, "Object write of %s completed successfully", mn->object);
|
| 1911 |
mio->copyups--; |
| 1912 |
signal_mapnode(mn); |
| 1913 |
signal_pr(pr); |
| 1914 |
} else if (req->op == X_COPY) { |
| 1915 |
// issue write_object;
|
| 1916 |
mn->flags &= ~MF_OBJECT_COPYING; |
| 1917 |
struct map *map = mn->map;
|
| 1918 |
if (!map){
|
| 1919 |
XSEGLOG2(&lc, E, "Object %s has not map back pointer", mn->object);
|
| 1920 |
goto out_err;
|
| 1921 |
} |
| 1922 |
|
| 1923 |
/* construct a tmp map_node for writing purposes */
|
| 1924 |
char *target = xseg_get_target(peer->xseg, req);
|
| 1925 |
struct map_node newmn = *mn;
|
| 1926 |
newmn.flags = MF_OBJECT_EXIST; |
| 1927 |
strncpy(newmn.object, target, req->targetlen); |
| 1928 |
newmn.object[req->targetlen] = 0;
|
| 1929 |
newmn.objectlen = req->targetlen; |
| 1930 |
newmn.objectidx = mn->objectidx; |
| 1931 |
struct xseg_request *xreq = object_write(peer, pr, map, &newmn);
|
| 1932 |
if (!xreq){
|
| 1933 |
XSEGLOG2(&lc, E, "Object write returned error for object %s"
|
| 1934 |
"\n\t of map %s [%llu]",
|
| 1935 |
mn->object, map->volume, (unsigned long long) mn->objectidx); |
| 1936 |
goto out_err;
|
| 1937 |
} |
| 1938 |
mn->flags |= MF_OBJECT_WRITING; |
| 1939 |
__set_copyup_node (mio, xreq, mn); |
| 1940 |
|
| 1941 |
XSEGLOG2(&lc, I, "Object %s copy up completed. Pending writing.", mn->object);
|
| 1942 |
} else {
|
| 1943 |
//wtf??
|
| 1944 |
; |
| 1945 |
} |
| 1946 |
|
| 1947 |
out:
|
| 1948 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 1949 |
return;
|
| 1950 |
|
| 1951 |
out_err:
|
| 1952 |
mio->copyups--; |
| 1953 |
XSEGLOG2(&lc, D, "Mio->copyups: %u", mio->copyups);
|
| 1954 |
mio->err = 1;
|
| 1955 |
if (mn)
|
| 1956 |
signal_mapnode(mn); |
| 1957 |
signal_pr(pr); |
| 1958 |
goto out;
|
| 1959 |
|
| 1960 |
} |
| 1961 |
|
| 1962 |
struct r2o {
|
| 1963 |
struct map_node *mn;
|
| 1964 |
uint64_t offset; |
| 1965 |
uint64_t size; |
| 1966 |
}; |
| 1967 |
|
| 1968 |
static int req2objs(struct peer_req *pr, struct map *map, int write) |
| 1969 |
{
|
| 1970 |
int r = 0; |
| 1971 |
struct peerd *peer = pr->peer;
|
| 1972 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 1973 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 1974 |
uint32_t nr_objs = calc_nr_obj(pr->req); |
| 1975 |
uint64_t size = sizeof(struct xseg_reply_map) + |
| 1976 |
nr_objs * sizeof(struct xseg_reply_map_scatterlist); |
| 1977 |
uint32_t idx, i; |
| 1978 |
uint64_t rem_size, obj_index, obj_offset, obj_size; |
| 1979 |
struct map_node *mn;
|
| 1980 |
mio->copyups = 0;
|
| 1981 |
XSEGLOG2(&lc, D, "Calculated %u nr_objs", nr_objs);
|
| 1982 |
|
| 1983 |
/* get map_nodes of request */
|
| 1984 |
struct r2o *mns = malloc(sizeof(struct r2o)*nr_objs); |
| 1985 |
if (!mns){
|
| 1986 |
XSEGLOG2(&lc, E, "Cannot allocate mns");
|
| 1987 |
return -1; |
| 1988 |
} |
| 1989 |
idx = 0;
|
| 1990 |
rem_size = pr->req->size; |
| 1991 |
obj_index = pr->req->offset / block_size; |
| 1992 |
obj_offset = pr->req->offset & (block_size -1); //modulo |
| 1993 |
obj_size = (obj_offset + rem_size > block_size) ? block_size - obj_offset : rem_size; |
| 1994 |
mn = get_mapnode(map, obj_index); |
| 1995 |
if (!mn) {
|
| 1996 |
XSEGLOG2(&lc, E, "Cannot find obj_index %llu\n", (unsigned long long) obj_index); |
| 1997 |
r = -1;
|
| 1998 |
goto out;
|
| 1999 |
} |
| 2000 |
mns[idx].mn = mn; |
| 2001 |
mns[idx].offset = obj_offset; |
| 2002 |
mns[idx].size = obj_size; |
| 2003 |
rem_size -= obj_size; |
| 2004 |
while (rem_size > 0) { |
| 2005 |
idx++; |
| 2006 |
obj_index++; |
| 2007 |
obj_offset = 0;
|
| 2008 |
obj_size = (rem_size > block_size) ? block_size : rem_size; |
| 2009 |
rem_size -= obj_size; |
| 2010 |
mn = get_mapnode(map, obj_index); |
| 2011 |
if (!mn) {
|
| 2012 |
XSEGLOG2(&lc, E, "Cannot find obj_index %llu\n", (unsigned long long) obj_index); |
| 2013 |
r = -1;
|
| 2014 |
goto out;
|
| 2015 |
} |
| 2016 |
mns[idx].mn = mn; |
| 2017 |
mns[idx].offset = obj_offset; |
| 2018 |
mns[idx].size = obj_size; |
| 2019 |
} |
| 2020 |
if (write) {
|
| 2021 |
int can_wait = 0; |
| 2022 |
mio->cb=copyup_cb; |
| 2023 |
/* do a first scan and issue as many copyups as we can.
|
| 2024 |
* then retry and wait when an object is not ready.
|
| 2025 |
* this could be done better, since now we wait also on the
|
| 2026 |
* pending copyups
|
| 2027 |
*/
|
| 2028 |
int j;
|
| 2029 |
for (j = 0; j < 2 && !mio->err; j++) { |
| 2030 |
for (i = 0; i < (idx+1); i++) { |
| 2031 |
mn = mns[i].mn; |
| 2032 |
//do copyups
|
| 2033 |
if (mn->flags & MF_OBJECT_NOT_READY){
|
| 2034 |
if (!can_wait)
|
| 2035 |
continue;
|
| 2036 |
wait_on_mapnode(mn, mn->flags & MF_OBJECT_NOT_READY); |
| 2037 |
if (mn->flags & MF_OBJECT_DESTROYED){
|
| 2038 |
mio->err = 1;
|
| 2039 |
continue;
|
| 2040 |
} |
| 2041 |
} |
| 2042 |
|
| 2043 |
if (!(mn->flags & MF_OBJECT_EXIST)) {
|
| 2044 |
//calc new_target, copy up object
|
| 2045 |
if (copyup_object(peer, mn, pr) == NULL){ |
| 2046 |
XSEGLOG2(&lc, E, "Error in copy up object");
|
| 2047 |
mio->err = 1;
|
| 2048 |
} else {
|
| 2049 |
mio->copyups++; |
| 2050 |
} |
| 2051 |
} |
| 2052 |
|
| 2053 |
if (mio->err){
|
| 2054 |
XSEGLOG2(&lc, E, "Mio-err, pending_copyups: %d", mio->copyups);
|
| 2055 |
break;
|
| 2056 |
} |
| 2057 |
} |
| 2058 |
can_wait = 1;
|
| 2059 |
} |
| 2060 |
wait_on_pr(pr, mio->copyups > 0);
|
| 2061 |
} |
| 2062 |
|
| 2063 |
if (mio->err){
|
| 2064 |
r = -1;
|
| 2065 |
XSEGLOG2(&lc, E, "Mio->err");
|
| 2066 |
goto out;
|
| 2067 |
} |
| 2068 |
|
| 2069 |
/* resize request to fit reply */
|
| 2070 |
char buf[XSEG_MAX_TARGETLEN];
|
| 2071 |
strncpy(buf, target, pr->req->targetlen); |
| 2072 |
r = xseg_resize_request(peer->xseg, pr->req, pr->req->targetlen, size); |
| 2073 |
if (r < 0) { |
| 2074 |
XSEGLOG2(&lc, E, "Cannot resize request");
|
| 2075 |
goto out;
|
| 2076 |
} |
| 2077 |
target = xseg_get_target(peer->xseg, pr->req); |
| 2078 |
strncpy(target, buf, pr->req->targetlen); |
| 2079 |
|
| 2080 |
/* structure reply */
|
| 2081 |
struct xseg_reply_map *reply = (struct xseg_reply_map *) xseg_get_data(peer->xseg, pr->req); |
| 2082 |
reply->cnt = nr_objs; |
| 2083 |
for (i = 0; i < (idx+1); i++) { |
| 2084 |
strncpy(reply->segs[i].target, mns[i].mn->object, mns[i].mn->objectlen); |
| 2085 |
reply->segs[i].targetlen = mns[i].mn->objectlen; |
| 2086 |
reply->segs[i].offset = mns[i].offset; |
| 2087 |
reply->segs[i].size = mns[i].size; |
| 2088 |
} |
| 2089 |
out:
|
| 2090 |
for (i = 0; i < idx; i++) { |
| 2091 |
put_mapnode(mns[i].mn); |
| 2092 |
} |
| 2093 |
free(mns); |
| 2094 |
mio->cb = NULL;
|
| 2095 |
return r;
|
| 2096 |
} |
| 2097 |
|
| 2098 |
static int do_dropcache(struct peer_req *pr, struct map *map) |
| 2099 |
{
|
| 2100 |
struct map_node *mn;
|
| 2101 |
struct peerd *peer = pr->peer;
|
| 2102 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 2103 |
uint64_t i; |
| 2104 |
XSEGLOG2(&lc, I, "Dropping cache for map %s", map->volume);
|
| 2105 |
map->flags |= MF_MAP_DROPPING_CACHE; |
| 2106 |
for (i = 0; i < calc_map_obj(map); i++) { |
| 2107 |
mn = get_mapnode(map, i); |
| 2108 |
if (mn) {
|
| 2109 |
if (!(mn->flags & MF_OBJECT_DESTROYED)){
|
| 2110 |
//make sure all pending operations on all objects are completed
|
| 2111 |
wait_on_mapnode(mn, mn->flags & MF_OBJECT_NOT_READY); |
| 2112 |
mn->flags |= MF_OBJECT_DESTROYED; |
| 2113 |
} |
| 2114 |
put_mapnode(mn); |
| 2115 |
} |
| 2116 |
} |
| 2117 |
map->flags &= ~MF_MAP_DROPPING_CACHE; |
| 2118 |
map->flags |= MF_MAP_DESTROYED; |
| 2119 |
remove_map(mapper, map); |
| 2120 |
XSEGLOG2(&lc, I, "Dropping cache for map %s completed", map->volume);
|
| 2121 |
put_map(map); // put map here to destroy it (matches m->ref = 1 on map create)
|
| 2122 |
return 0; |
| 2123 |
} |
| 2124 |
|
| 2125 |
static int do_info(struct peer_req *pr, struct map *map) |
| 2126 |
{
|
| 2127 |
struct peerd *peer = pr->peer;
|
| 2128 |
struct xseg_reply_info *xinfo = (struct xseg_reply_info *) xseg_get_data(peer->xseg, pr->req); |
| 2129 |
xinfo->size = map->size; |
| 2130 |
return 0; |
| 2131 |
} |
| 2132 |
|
| 2133 |
|
| 2134 |
static int do_open(struct peer_req *pr, struct map *map) |
| 2135 |
{
|
| 2136 |
if (map->flags & MF_MAP_EXCLUSIVE){
|
| 2137 |
return 0; |
| 2138 |
} |
| 2139 |
else {
|
| 2140 |
return -1; |
| 2141 |
} |
| 2142 |
} |
| 2143 |
|
| 2144 |
static int do_close(struct peer_req *pr, struct map *map) |
| 2145 |
{
|
| 2146 |
if (map->flags & MF_MAP_EXCLUSIVE){
|
| 2147 |
/* do not drop cache if close failed and map not deleted */
|
| 2148 |
if (close_map(pr, map) < 0 && !(map->flags & MF_MAP_DELETED)) |
| 2149 |
return -1; |
| 2150 |
} |
| 2151 |
return do_dropcache(pr, map);
|
| 2152 |
} |
| 2153 |
|
| 2154 |
static int do_snapshot(struct peer_req *pr, struct map *map) |
| 2155 |
{
|
| 2156 |
uint64_t i; |
| 2157 |
struct peerd *peer = pr->peer;
|
| 2158 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 2159 |
struct map_node *mn;
|
| 2160 |
struct xseg_request *req;
|
| 2161 |
|
| 2162 |
if (!(map->flags & MF_MAP_EXCLUSIVE)){
|
| 2163 |
XSEGLOG2(&lc, E, "Map was not opened exclusively");
|
| 2164 |
return -1; |
| 2165 |
} |
| 2166 |
XSEGLOG2(&lc, I, "Starting snapshot for map %s", map->volume);
|
| 2167 |
map->flags |= MF_MAP_SNAPSHOTTING; |
| 2168 |
|
| 2169 |
uint64_t nr_obj = calc_map_obj(map); |
| 2170 |
mio->cb = snapshot_cb; |
| 2171 |
mio->snap_pending = 0;
|
| 2172 |
mio->err = 0;
|
| 2173 |
for (i = 0; i < nr_obj; i++){ |
| 2174 |
|
| 2175 |
/* throttle pending snapshots
|
| 2176 |
* this should be nr_ops of the blocker, but since we don't know
|
| 2177 |
* that, we assume based on our own nr_ops
|
| 2178 |
*/
|
| 2179 |
wait_on_pr(pr, mio->snap_pending >= peer->nr_ops); |
| 2180 |
|
| 2181 |
mn = get_mapnode(map, i); |
| 2182 |
if (!mn)
|
| 2183 |
//warning?
|
| 2184 |
continue;
|
| 2185 |
if (!(mn->flags & MF_OBJECT_EXIST)){
|
| 2186 |
put_mapnode(mn); |
| 2187 |
continue;
|
| 2188 |
} |
| 2189 |
// make sure all pending operations on all objects are completed
|
| 2190 |
wait_on_mapnode(mn, mn->flags & MF_OBJECT_NOT_READY); |
| 2191 |
|
| 2192 |
/* TODO will this ever happen?? */
|
| 2193 |
if (mn->flags & MF_OBJECT_DESTROYED){
|
| 2194 |
put_mapnode(mn); |
| 2195 |
continue;
|
| 2196 |
} |
| 2197 |
|
| 2198 |
req = __snapshot_object(pr, mn); |
| 2199 |
if (!req){
|
| 2200 |
mio->err = 1;
|
| 2201 |
put_mapnode(mn); |
| 2202 |
break;
|
| 2203 |
} |
| 2204 |
mio->snap_pending++; |
| 2205 |
/* do not put_mapnode here. cb does that */
|
| 2206 |
} |
| 2207 |
|
| 2208 |
wait_on_pr(pr, mio->snap_pending > 0);
|
| 2209 |
mio->cb = NULL;
|
| 2210 |
|
| 2211 |
if (mio->err)
|
| 2212 |
goto out_err;
|
| 2213 |
|
| 2214 |
/* calculate name of snapshot */
|
| 2215 |
struct map tmp_map = *map;
|
| 2216 |
unsigned char sha[SHA256_DIGEST_SIZE]; |
| 2217 |
unsigned char *buf = malloc(block_size); |
| 2218 |
char newvolumename[MAX_VOLUME_LEN];
|
| 2219 |
uint32_t newvolumenamelen = HEXLIFIED_SHA256_DIGEST_SIZE; |
| 2220 |
uint64_t pos = 0;
|
| 2221 |
uint64_t max_objidx = calc_map_obj(map); |
| 2222 |
int r;
|
| 2223 |
|
| 2224 |
for (i = 0; i < max_objidx; i++) { |
| 2225 |
mn = find_object(map, i); |
| 2226 |
if (!mn){
|
| 2227 |
XSEGLOG2(&lc, E, "Cannot find object %llu for map %s",
|
| 2228 |
(unsigned long long) i, map->volume); |
| 2229 |
goto out_err;
|
| 2230 |
} |
| 2231 |
v0_object_to_map(mn, buf+pos); |
| 2232 |
pos += v0_objectsize_in_map; |
| 2233 |
} |
| 2234 |
// SHA256(buf, pos, sha);
|
| 2235 |
merkle_hash(buf, pos, sha); |
| 2236 |
hexlify(sha, newvolumename); |
| 2237 |
strncpy(tmp_map.volume, newvolumename, newvolumenamelen); |
| 2238 |
tmp_map.volumelen = newvolumenamelen; |
| 2239 |
free(buf); |
| 2240 |
tmp_map.version = 0; // set volume version to pithos image |
| 2241 |
|
| 2242 |
/* write the map of the Snapshot */
|
| 2243 |
r = write_map(pr, &tmp_map); |
| 2244 |
if (r < 0) |
| 2245 |
goto out_err;
|
| 2246 |
char targetbuf[XSEG_MAX_TARGETLEN];
|
| 2247 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2248 |
strncpy(targetbuf, target, pr->req->targetlen); |
| 2249 |
r = xseg_resize_request(peer->xseg, pr->req, pr->req->targetlen, |
| 2250 |
sizeof(struct xseg_reply_snapshot)); |
| 2251 |
if (r < 0){ |
| 2252 |
XSEGLOG2(&lc, E, "Cannot resize request");
|
| 2253 |
goto out_err;
|
| 2254 |
} |
| 2255 |
target = xseg_get_target(peer->xseg, pr->req); |
| 2256 |
strncpy(target, targetbuf, pr->req->targetlen); |
| 2257 |
|
| 2258 |
struct xseg_reply_snapshot *xreply = (struct xseg_reply_snapshot *) |
| 2259 |
xseg_get_data(peer->xseg, pr->req); |
| 2260 |
strncpy(xreply->target, newvolumename, newvolumenamelen); |
| 2261 |
xreply->targetlen = newvolumenamelen; |
| 2262 |
map->flags &= ~MF_MAP_SNAPSHOTTING; |
| 2263 |
XSEGLOG2(&lc, I, "Snapshot for map %s completed", map->volume);
|
| 2264 |
return 0; |
| 2265 |
|
| 2266 |
out_err:
|
| 2267 |
map->flags &= ~MF_MAP_SNAPSHOTTING; |
| 2268 |
XSEGLOG2(&lc, E, "Snapshot for map %s failed", map->volume);
|
| 2269 |
return -1; |
| 2270 |
} |
| 2271 |
|
| 2272 |
|
| 2273 |
static int do_destroy(struct peer_req *pr, struct map *map) |
| 2274 |
{
|
| 2275 |
uint64_t i; |
| 2276 |
struct peerd *peer = pr->peer;
|
| 2277 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 2278 |
struct map_node *mn;
|
| 2279 |
struct xseg_request *req;
|
| 2280 |
|
| 2281 |
if (!(map->flags & MF_MAP_EXCLUSIVE))
|
| 2282 |
return -1; |
| 2283 |
|
| 2284 |
XSEGLOG2(&lc, I, "Destroying map %s", map->volume);
|
| 2285 |
req = __delete_map(pr, map); |
| 2286 |
if (!req)
|
| 2287 |
return -1; |
| 2288 |
wait_on_pr(pr, (!((req->state & XS_FAILED)||(req->state & XS_SERVED)))); |
| 2289 |
if (req->state & XS_FAILED){
|
| 2290 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 2291 |
map->flags &= ~MF_MAP_DELETING; |
| 2292 |
return -1; |
| 2293 |
} |
| 2294 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 2295 |
|
| 2296 |
uint64_t nr_obj = calc_map_obj(map); |
| 2297 |
mio->cb = deletion_cb; |
| 2298 |
mio->del_pending = 0;
|
| 2299 |
mio->err = 0;
|
| 2300 |
for (i = 0; i < nr_obj; i++){ |
| 2301 |
|
| 2302 |
/* throttle pending deletions
|
| 2303 |
* this should be nr_ops of the blocker, but since we don't know
|
| 2304 |
* that, we assume based on our own nr_ops
|
| 2305 |
*/
|
| 2306 |
wait_on_pr(pr, mio->del_pending >= peer->nr_ops); |
| 2307 |
|
| 2308 |
mn = get_mapnode(map, i); |
| 2309 |
if (!mn)
|
| 2310 |
continue;
|
| 2311 |
if (mn->flags & MF_OBJECT_DESTROYED){
|
| 2312 |
put_mapnode(mn); |
| 2313 |
continue;
|
| 2314 |
} |
| 2315 |
if (!(mn->flags & MF_OBJECT_EXIST)){
|
| 2316 |
mn->flags |= MF_OBJECT_DESTROYED; |
| 2317 |
put_mapnode(mn); |
| 2318 |
continue;
|
| 2319 |
} |
| 2320 |
|
| 2321 |
// make sure all pending operations on all objects are completed
|
| 2322 |
wait_on_mapnode(mn, mn->flags & MF_OBJECT_NOT_READY); |
| 2323 |
|
| 2324 |
req = __delete_object(pr, mn); |
| 2325 |
if (!req){
|
| 2326 |
mio->err = 1;
|
| 2327 |
put_mapnode(mn); |
| 2328 |
continue;
|
| 2329 |
} |
| 2330 |
mio->del_pending++; |
| 2331 |
/* do not put_mapnode here. cb does that */
|
| 2332 |
} |
| 2333 |
|
| 2334 |
wait_on_pr(pr, mio->del_pending > 0);
|
| 2335 |
|
| 2336 |
mio->cb = NULL;
|
| 2337 |
map->flags &= ~MF_MAP_DELETING; |
| 2338 |
map->flags |= MF_MAP_DELETED; |
| 2339 |
XSEGLOG2(&lc, I, "Destroyed map %s", map->volume);
|
| 2340 |
return do_close(pr, map);
|
| 2341 |
} |
| 2342 |
|
| 2343 |
static int do_mapr(struct peer_req *pr, struct map *map) |
| 2344 |
{
|
| 2345 |
struct peerd *peer = pr->peer;
|
| 2346 |
int r = req2objs(pr, map, 0); |
| 2347 |
if (r < 0){ |
| 2348 |
XSEGLOG2(&lc, I, "Map r of map %s, range: %llu-%llu failed",
|
| 2349 |
map->volume, |
| 2350 |
(unsigned long long) pr->req->offset, |
| 2351 |
(unsigned long long) (pr->req->offset + pr->req->size)); |
| 2352 |
return -1; |
| 2353 |
} |
| 2354 |
XSEGLOG2(&lc, I, "Map r of map %s, range: %llu-%llu completed",
|
| 2355 |
map->volume, |
| 2356 |
(unsigned long long) pr->req->offset, |
| 2357 |
(unsigned long long) (pr->req->offset + pr->req->size)); |
| 2358 |
XSEGLOG2(&lc, D, "Req->offset: %llu, req->size: %llu",
|
| 2359 |
(unsigned long long) pr->req->offset, |
| 2360 |
(unsigned long long) pr->req->size); |
| 2361 |
char buf[XSEG_MAX_TARGETLEN+1]; |
| 2362 |
struct xseg_reply_map *reply = (struct xseg_reply_map *) xseg_get_data(peer->xseg, pr->req); |
| 2363 |
int i;
|
| 2364 |
for (i = 0; i < reply->cnt; i++) { |
| 2365 |
XSEGLOG2(&lc, D, "i: %d, reply->cnt: %u",i, reply->cnt);
|
| 2366 |
strncpy(buf, reply->segs[i].target, reply->segs[i].targetlen); |
| 2367 |
buf[reply->segs[i].targetlen] = 0;
|
| 2368 |
XSEGLOG2(&lc, D, "%d: Object: %s, offset: %llu, size: %llu", i, buf,
|
| 2369 |
(unsigned long long) reply->segs[i].offset, |
| 2370 |
(unsigned long long) reply->segs[i].size); |
| 2371 |
} |
| 2372 |
return 0; |
| 2373 |
} |
| 2374 |
|
| 2375 |
static int do_mapw(struct peer_req *pr, struct map *map) |
| 2376 |
{
|
| 2377 |
struct peerd *peer = pr->peer;
|
| 2378 |
int r = req2objs(pr, map, 1); |
| 2379 |
if (r < 0){ |
| 2380 |
XSEGLOG2(&lc, I, "Map w of map %s, range: %llu-%llu failed",
|
| 2381 |
map->volume, |
| 2382 |
(unsigned long long) pr->req->offset, |
| 2383 |
(unsigned long long) (pr->req->offset + pr->req->size)); |
| 2384 |
return -1; |
| 2385 |
} |
| 2386 |
XSEGLOG2(&lc, I, "Map w of map %s, range: %llu-%llu completed",
|
| 2387 |
map->volume, |
| 2388 |
(unsigned long long) pr->req->offset, |
| 2389 |
(unsigned long long) (pr->req->offset + pr->req->size)); |
| 2390 |
XSEGLOG2(&lc, D, "Req->offset: %llu, req->size: %llu",
|
| 2391 |
(unsigned long long) pr->req->offset, |
| 2392 |
(unsigned long long) pr->req->size); |
| 2393 |
char buf[XSEG_MAX_TARGETLEN+1]; |
| 2394 |
struct xseg_reply_map *reply = (struct xseg_reply_map *) xseg_get_data(peer->xseg, pr->req); |
| 2395 |
int i;
|
| 2396 |
for (i = 0; i < reply->cnt; i++) { |
| 2397 |
XSEGLOG2(&lc, D, "i: %d, reply->cnt: %u",i, reply->cnt);
|
| 2398 |
strncpy(buf, reply->segs[i].target, reply->segs[i].targetlen); |
| 2399 |
buf[reply->segs[i].targetlen] = 0;
|
| 2400 |
XSEGLOG2(&lc, D, "%d: Object: %s, offset: %llu, size: %llu", i, buf,
|
| 2401 |
(unsigned long long) reply->segs[i].offset, |
| 2402 |
(unsigned long long) reply->segs[i].size); |
| 2403 |
} |
| 2404 |
return 0; |
| 2405 |
} |
| 2406 |
|
| 2407 |
//here map is the parent map
|
| 2408 |
static int do_clone(struct peer_req *pr, struct map *map) |
| 2409 |
{
|
| 2410 |
int r;
|
| 2411 |
struct peerd *peer = pr->peer;
|
| 2412 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 2413 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2414 |
struct map *clonemap;
|
| 2415 |
struct xseg_request_clone *xclone =
|
| 2416 |
(struct xseg_request_clone *) xseg_get_data(peer->xseg, pr->req);
|
| 2417 |
|
| 2418 |
XSEGLOG2(&lc, I, "Cloning map %s", map->volume);
|
| 2419 |
|
| 2420 |
clonemap = create_map(mapper, target, pr->req->targetlen, MF_ARCHIP); |
| 2421 |
if (!clonemap)
|
| 2422 |
return -1; |
| 2423 |
|
| 2424 |
/* open map to get exclusive access to map */
|
| 2425 |
r = open_map(pr, clonemap, 0);
|
| 2426 |
if (r < 0){ |
| 2427 |
XSEGLOG2(&lc, E, "Cannot open map %s", clonemap->volume);
|
| 2428 |
XSEGLOG2(&lc, E, "Target volume %s exists", clonemap->volume);
|
| 2429 |
goto out_err;
|
| 2430 |
} |
| 2431 |
r = load_map(pr, clonemap); |
| 2432 |
if (r >= 0) { |
| 2433 |
XSEGLOG2(&lc, E, "Target volume %s exists", clonemap->volume);
|
| 2434 |
goto out_err;
|
| 2435 |
} |
| 2436 |
|
| 2437 |
if (xclone->size == -1) |
| 2438 |
clonemap->size = map->size; |
| 2439 |
else
|
| 2440 |
clonemap->size = xclone->size; |
| 2441 |
if (clonemap->size < map->size){
|
| 2442 |
XSEGLOG2(&lc, W, "Requested clone size (%llu) < map size (%llu)"
|
| 2443 |
"\n\t for requested clone %s",
|
| 2444 |
(unsigned long long) xclone->size, |
| 2445 |
(unsigned long long) map->size, clonemap->volume); |
| 2446 |
goto out_err;
|
| 2447 |
} |
| 2448 |
if (clonemap->size > MAX_VOLUME_SIZE) {
|
| 2449 |
XSEGLOG2(&lc, E, "Requested size %llu > max volume size %llu"
|
| 2450 |
"\n\t for volume %s",
|
| 2451 |
clonemap->size, MAX_VOLUME_SIZE, clonemap->volume); |
| 2452 |
goto out_err;
|
| 2453 |
} |
| 2454 |
|
| 2455 |
//alloc and init map_nodes
|
| 2456 |
//unsigned long c = clonemap->size/block_size + 1;
|
| 2457 |
unsigned long c = calc_map_obj(clonemap); |
| 2458 |
struct map_node *map_nodes = calloc(c, sizeof(struct map_node)); |
| 2459 |
if (!map_nodes){
|
| 2460 |
goto out_err;
|
| 2461 |
} |
| 2462 |
int i;
|
| 2463 |
//for (i = 0; i < clonemap->size/block_size + 1; i++) {
|
| 2464 |
for (i = 0; i < c; i++) { |
| 2465 |
struct map_node *mn = get_mapnode(map, i);
|
| 2466 |
if (mn) {
|
| 2467 |
strncpy(map_nodes[i].object, mn->object, mn->objectlen); |
| 2468 |
map_nodes[i].objectlen = mn->objectlen; |
| 2469 |
put_mapnode(mn); |
| 2470 |
} else {
|
| 2471 |
strncpy(map_nodes[i].object, zero_block, ZERO_BLOCK_LEN); |
| 2472 |
map_nodes[i].objectlen = ZERO_BLOCK_LEN; |
| 2473 |
} |
| 2474 |
map_nodes[i].object[map_nodes[i].objectlen] = 0; //NULL terminate |
| 2475 |
map_nodes[i].flags = 0;
|
| 2476 |
map_nodes[i].objectidx = i; |
| 2477 |
map_nodes[i].map = clonemap; |
| 2478 |
map_nodes[i].ref = 1;
|
| 2479 |
map_nodes[i].waiters = 0;
|
| 2480 |
map_nodes[i].cond = st_cond_new(); //FIXME errcheck;
|
| 2481 |
r = insert_object(clonemap, &map_nodes[i]); |
| 2482 |
if (r < 0){ |
| 2483 |
XSEGLOG2(&lc, E, "Cannot insert object %d to map %s", i, clonemap->volume);
|
| 2484 |
goto out_err;
|
| 2485 |
} |
| 2486 |
} |
| 2487 |
|
| 2488 |
r = write_map(pr, clonemap); |
| 2489 |
if (r < 0){ |
| 2490 |
XSEGLOG2(&lc, E, "Cannot write map %s", clonemap->volume);
|
| 2491 |
goto out_err;
|
| 2492 |
} |
| 2493 |
do_close(pr, clonemap); |
| 2494 |
return 0; |
| 2495 |
|
| 2496 |
out_err:
|
| 2497 |
do_close(pr, clonemap); |
| 2498 |
return -1; |
| 2499 |
} |
| 2500 |
|
| 2501 |
static int open_load_map(struct peer_req *pr, struct map *map, uint32_t flags) |
| 2502 |
{
|
| 2503 |
int r, opened = 0; |
| 2504 |
if (flags & MF_EXCLUSIVE){
|
| 2505 |
r = open_map(pr, map, flags); |
| 2506 |
if (r < 0) { |
| 2507 |
if (flags & MF_FORCE){
|
| 2508 |
return -1; |
| 2509 |
} |
| 2510 |
} else {
|
| 2511 |
opened = 1;
|
| 2512 |
} |
| 2513 |
} |
| 2514 |
r = load_map(pr, map); |
| 2515 |
if (r < 0 && opened){ |
| 2516 |
close_map(pr, map); |
| 2517 |
} |
| 2518 |
return r;
|
| 2519 |
} |
| 2520 |
|
| 2521 |
struct map * get_map(struct peer_req *pr, char *name, uint32_t namelen, |
| 2522 |
uint32_t flags) |
| 2523 |
{
|
| 2524 |
int r;
|
| 2525 |
struct peerd *peer = pr->peer;
|
| 2526 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 2527 |
struct map *map = find_map_len(mapper, name, namelen, flags);
|
| 2528 |
if (!map){
|
| 2529 |
if (flags & MF_LOAD){
|
| 2530 |
map = create_map(mapper, name, namelen, flags); |
| 2531 |
if (!map)
|
| 2532 |
return NULL; |
| 2533 |
r = open_load_map(pr, map, flags); |
| 2534 |
if (r < 0){ |
| 2535 |
do_dropcache(pr, map); |
| 2536 |
return NULL; |
| 2537 |
} |
| 2538 |
} else {
|
| 2539 |
return NULL; |
| 2540 |
} |
| 2541 |
} else if (map->flags & MF_MAP_DESTROYED){ |
| 2542 |
return NULL; |
| 2543 |
} |
| 2544 |
__get_map(map); |
| 2545 |
return map;
|
| 2546 |
|
| 2547 |
} |
| 2548 |
|
| 2549 |
static int map_action(int (action)(struct peer_req *pr, struct map *map), |
| 2550 |
struct peer_req *pr, char *name, uint32_t namelen, uint32_t flags) |
| 2551 |
{
|
| 2552 |
//struct peerd *peer = pr->peer;
|
| 2553 |
struct map *map;
|
| 2554 |
start:
|
| 2555 |
map = get_map(pr, name, namelen, flags); |
| 2556 |
if (!map)
|
| 2557 |
return -1; |
| 2558 |
if (map->flags & MF_MAP_NOT_READY){
|
| 2559 |
wait_on_map(map, (map->flags & MF_MAP_NOT_READY)); |
| 2560 |
put_map(map); |
| 2561 |
goto start;
|
| 2562 |
} |
| 2563 |
int r = action(pr, map);
|
| 2564 |
//always drop cache if map not read exclusively
|
| 2565 |
if (!(map->flags & MF_MAP_EXCLUSIVE))
|
| 2566 |
do_dropcache(pr, map); |
| 2567 |
signal_map(map); |
| 2568 |
put_map(map); |
| 2569 |
return r;
|
| 2570 |
} |
| 2571 |
|
| 2572 |
void * handle_info(struct peer_req *pr) |
| 2573 |
{
|
| 2574 |
struct peerd *peer = pr->peer;
|
| 2575 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2576 |
int r = map_action(do_info, pr, target, pr->req->targetlen,
|
| 2577 |
MF_ARCHIP|MF_LOAD); |
| 2578 |
if (r < 0) |
| 2579 |
fail(peer, pr); |
| 2580 |
else
|
| 2581 |
complete(peer, pr); |
| 2582 |
ta--; |
| 2583 |
return NULL; |
| 2584 |
} |
| 2585 |
|
| 2586 |
void * handle_clone(struct peer_req *pr) |
| 2587 |
{
|
| 2588 |
int r;
|
| 2589 |
struct peerd *peer = pr->peer;
|
| 2590 |
struct xseg_request_clone *xclone = (struct xseg_request_clone *) xseg_get_data(peer->xseg, pr->req); |
| 2591 |
if (!xclone) {
|
| 2592 |
r = -1;
|
| 2593 |
goto out;
|
| 2594 |
} |
| 2595 |
|
| 2596 |
if (xclone->targetlen){
|
| 2597 |
/* if snap was defined */
|
| 2598 |
//support clone only from pithos
|
| 2599 |
r = map_action(do_clone, pr, xclone->target, xclone->targetlen, |
| 2600 |
MF_LOAD); |
| 2601 |
} else {
|
| 2602 |
/* else try to create a new volume */
|
| 2603 |
XSEGLOG2(&lc, I, "Creating volume");
|
| 2604 |
if (!xclone->size){
|
| 2605 |
XSEGLOG2(&lc, E, "Cannot create volume. Size not specified");
|
| 2606 |
r = -1;
|
| 2607 |
goto out;
|
| 2608 |
} |
| 2609 |
if (xclone->size > MAX_VOLUME_SIZE) {
|
| 2610 |
XSEGLOG2(&lc, E, "Requested size %llu > max volume "
|
| 2611 |
"size %llu", xclone->size, MAX_VOLUME_SIZE);
|
| 2612 |
r = -1;
|
| 2613 |
goto out;
|
| 2614 |
} |
| 2615 |
|
| 2616 |
struct map *map;
|
| 2617 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2618 |
|
| 2619 |
//create a new empty map of size
|
| 2620 |
map = create_map(mapper, target, pr->req->targetlen, MF_ARCHIP); |
| 2621 |
if (!map){
|
| 2622 |
r = -1;
|
| 2623 |
goto out;
|
| 2624 |
} |
| 2625 |
/* open map to get exclusive access to map */
|
| 2626 |
r = open_map(pr, map, 0);
|
| 2627 |
if (r < 0){ |
| 2628 |
XSEGLOG2(&lc, E, "Cannot open map %s", map->volume);
|
| 2629 |
XSEGLOG2(&lc, E, "Target volume %s exists", map->volume);
|
| 2630 |
do_dropcache(pr, map); |
| 2631 |
r = -1;
|
| 2632 |
goto out;
|
| 2633 |
} |
| 2634 |
r = load_map(pr, map); |
| 2635 |
if (r >= 0) { |
| 2636 |
XSEGLOG2(&lc, E, "Map exists %s", map->volume);
|
| 2637 |
do_close(pr, map); |
| 2638 |
r = -1;
|
| 2639 |
goto out;
|
| 2640 |
} |
| 2641 |
map->size = xclone->size; |
| 2642 |
//populate_map with zero objects;
|
| 2643 |
uint64_t nr_objs = xclone->size / block_size; |
| 2644 |
if (xclone->size % block_size)
|
| 2645 |
nr_objs++; |
| 2646 |
|
| 2647 |
struct map_node *map_nodes = calloc(nr_objs, sizeof(struct map_node)); |
| 2648 |
if (!map_nodes){
|
| 2649 |
do_close(pr, map); |
| 2650 |
r = -1;
|
| 2651 |
goto out;
|
| 2652 |
} |
| 2653 |
|
| 2654 |
uint64_t i; |
| 2655 |
for (i = 0; i < nr_objs; i++) { |
| 2656 |
strncpy(map_nodes[i].object, zero_block, ZERO_BLOCK_LEN); |
| 2657 |
map_nodes[i].objectlen = ZERO_BLOCK_LEN; |
| 2658 |
map_nodes[i].object[map_nodes[i].objectlen] = 0; //NULL terminate |
| 2659 |
map_nodes[i].flags = 0;
|
| 2660 |
map_nodes[i].objectidx = i; |
| 2661 |
map_nodes[i].map = map; |
| 2662 |
map_nodes[i].ref = 1;
|
| 2663 |
map_nodes[i].waiters = 0;
|
| 2664 |
map_nodes[i].cond = st_cond_new(); //FIXME errcheck;
|
| 2665 |
r = insert_object(map, &map_nodes[i]); |
| 2666 |
if (r < 0){ |
| 2667 |
do_close(pr, map); |
| 2668 |
r = -1;
|
| 2669 |
goto out;
|
| 2670 |
} |
| 2671 |
} |
| 2672 |
r = write_map(pr, map); |
| 2673 |
if (r < 0){ |
| 2674 |
XSEGLOG2(&lc, E, "Cannot write map %s", map->volume);
|
| 2675 |
do_close(pr, map); |
| 2676 |
goto out;
|
| 2677 |
} |
| 2678 |
XSEGLOG2(&lc, I, "Volume %s created", map->volume);
|
| 2679 |
r = 0;
|
| 2680 |
do_close(pr, map); //drop cache here for consistency
|
| 2681 |
} |
| 2682 |
out:
|
| 2683 |
if (r < 0) |
| 2684 |
fail(peer, pr); |
| 2685 |
else
|
| 2686 |
complete(peer, pr); |
| 2687 |
ta--; |
| 2688 |
return NULL; |
| 2689 |
} |
| 2690 |
|
| 2691 |
void * handle_mapr(struct peer_req *pr) |
| 2692 |
{
|
| 2693 |
struct peerd *peer = pr->peer;
|
| 2694 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2695 |
int r = map_action(do_mapr, pr, target, pr->req->targetlen,
|
| 2696 |
MF_ARCHIP|MF_LOAD|MF_EXCLUSIVE); |
| 2697 |
if (r < 0) |
| 2698 |
fail(peer, pr); |
| 2699 |
else
|
| 2700 |
complete(peer, pr); |
| 2701 |
ta--; |
| 2702 |
return NULL; |
| 2703 |
} |
| 2704 |
|
| 2705 |
void * handle_mapw(struct peer_req *pr) |
| 2706 |
{
|
| 2707 |
struct peerd *peer = pr->peer;
|
| 2708 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2709 |
int r = map_action(do_mapw, pr, target, pr->req->targetlen,
|
| 2710 |
MF_ARCHIP|MF_LOAD|MF_EXCLUSIVE|MF_FORCE); |
| 2711 |
if (r < 0) |
| 2712 |
fail(peer, pr); |
| 2713 |
else
|
| 2714 |
complete(peer, pr); |
| 2715 |
XSEGLOG2(&lc, D, "Ta: %d", ta);
|
| 2716 |
ta--; |
| 2717 |
return NULL; |
| 2718 |
} |
| 2719 |
|
| 2720 |
void * handle_destroy(struct peer_req *pr) |
| 2721 |
{
|
| 2722 |
struct peerd *peer = pr->peer;
|
| 2723 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2724 |
/* request EXCLUSIVE access, but do not force it.
|
| 2725 |
* check if succeeded on do_destroy
|
| 2726 |
*/
|
| 2727 |
int r = map_action(do_destroy, pr, target, pr->req->targetlen,
|
| 2728 |
MF_ARCHIP|MF_LOAD|MF_EXCLUSIVE); |
| 2729 |
if (r < 0) |
| 2730 |
fail(peer, pr); |
| 2731 |
else
|
| 2732 |
complete(peer, pr); |
| 2733 |
ta--; |
| 2734 |
return NULL; |
| 2735 |
} |
| 2736 |
|
| 2737 |
void * handle_open(struct peer_req *pr) |
| 2738 |
{
|
| 2739 |
struct peerd *peer = pr->peer;
|
| 2740 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2741 |
//here we do not want to load
|
| 2742 |
int r = map_action(do_open, pr, target, pr->req->targetlen,
|
| 2743 |
MF_ARCHIP|MF_LOAD|MF_EXCLUSIVE); |
| 2744 |
if (r < 0) |
| 2745 |
fail(peer, pr); |
| 2746 |
else
|
| 2747 |
complete(peer, pr); |
| 2748 |
ta--; |
| 2749 |
return NULL; |
| 2750 |
} |
| 2751 |
|
| 2752 |
void * handle_close(struct peer_req *pr) |
| 2753 |
{
|
| 2754 |
struct peerd *peer = pr->peer;
|
| 2755 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2756 |
//here we do not want to load
|
| 2757 |
int r = map_action(do_close, pr, target, pr->req->targetlen,
|
| 2758 |
MF_ARCHIP|MF_EXCLUSIVE|MF_FORCE); |
| 2759 |
if (r < 0) |
| 2760 |
fail(peer, pr); |
| 2761 |
else
|
| 2762 |
complete(peer, pr); |
| 2763 |
ta--; |
| 2764 |
return NULL; |
| 2765 |
} |
| 2766 |
|
| 2767 |
void * handle_snapshot(struct peer_req *pr) |
| 2768 |
{
|
| 2769 |
struct peerd *peer = pr->peer;
|
| 2770 |
char *target = xseg_get_target(peer->xseg, pr->req);
|
| 2771 |
/* request EXCLUSIVE access, but do not force it.
|
| 2772 |
* check if succeeded on do_snapshot
|
| 2773 |
*/
|
| 2774 |
int r = map_action(do_snapshot, pr, target, pr->req->targetlen,
|
| 2775 |
MF_ARCHIP|MF_LOAD|MF_EXCLUSIVE); |
| 2776 |
if (r < 0) |
| 2777 |
fail(peer, pr); |
| 2778 |
else
|
| 2779 |
complete(peer, pr); |
| 2780 |
ta--; |
| 2781 |
return NULL; |
| 2782 |
} |
| 2783 |
|
| 2784 |
int dispatch_accepted(struct peerd *peer, struct peer_req *pr, |
| 2785 |
struct xseg_request *req)
|
| 2786 |
{
|
| 2787 |
//struct mapperd *mapper = __get_mapperd(peer);
|
| 2788 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 2789 |
void *(*action)(struct peer_req *) = NULL; |
| 2790 |
|
| 2791 |
mio->state = ACCEPTED; |
| 2792 |
mio->err = 0;
|
| 2793 |
mio->cb = NULL;
|
| 2794 |
switch (pr->req->op) {
|
| 2795 |
/* primary xseg operations of mapper */
|
| 2796 |
case X_CLONE: action = handle_clone; break; |
| 2797 |
case X_MAPR: action = handle_mapr; break; |
| 2798 |
case X_MAPW: action = handle_mapw; break; |
| 2799 |
case X_SNAPSHOT: action = handle_snapshot; break; |
| 2800 |
case X_INFO: action = handle_info; break; |
| 2801 |
case X_DELETE: action = handle_destroy; break; |
| 2802 |
case X_OPEN: action = handle_open; break; |
| 2803 |
case X_CLOSE: action = handle_close; break; |
| 2804 |
default: fprintf(stderr, "mydispatch: unknown up\n"); break; |
| 2805 |
} |
| 2806 |
if (action){
|
| 2807 |
ta++; |
| 2808 |
mio->active = 1;
|
| 2809 |
st_thread_create(action, pr, 0, 0); |
| 2810 |
} |
| 2811 |
return 0; |
| 2812 |
|
| 2813 |
} |
| 2814 |
|
| 2815 |
int dispatch(struct peerd *peer, struct peer_req *pr, struct xseg_request *req, |
| 2816 |
enum dispatch_reason reason)
|
| 2817 |
{
|
| 2818 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 2819 |
(void) mapper;
|
| 2820 |
struct mapper_io *mio = __get_mapper_io(pr);
|
| 2821 |
(void) mio;
|
| 2822 |
|
| 2823 |
|
| 2824 |
if (reason == dispatch_accept)
|
| 2825 |
dispatch_accepted(peer, pr, req); |
| 2826 |
else {
|
| 2827 |
if (mio->cb){
|
| 2828 |
mio->cb(pr, req); |
| 2829 |
} else {
|
| 2830 |
signal_pr(pr); |
| 2831 |
} |
| 2832 |
} |
| 2833 |
return 0; |
| 2834 |
} |
| 2835 |
|
| 2836 |
int custom_peer_init(struct peerd *peer, int argc, char *argv[]) |
| 2837 |
{
|
| 2838 |
int i;
|
| 2839 |
|
| 2840 |
//FIXME error checks
|
| 2841 |
struct mapperd *mapperd = malloc(sizeof(struct mapperd)); |
| 2842 |
peer->priv = mapperd; |
| 2843 |
mapper = mapperd; |
| 2844 |
mapper->hashmaps = xhash_new(3, 0, STRING); |
| 2845 |
|
| 2846 |
for (i = 0; i < peer->nr_ops; i++) { |
| 2847 |
struct mapper_io *mio = malloc(sizeof(struct mapper_io)); |
| 2848 |
mio->copyups_nodes = xhash_new(3, 0, INTEGER); |
| 2849 |
mio->copyups = 0;
|
| 2850 |
mio->err = 0;
|
| 2851 |
mio->active = 0;
|
| 2852 |
peer->peer_reqs[i].priv = mio; |
| 2853 |
} |
| 2854 |
|
| 2855 |
mapper->bportno = -1;
|
| 2856 |
mapper->mbportno = -1;
|
| 2857 |
BEGIN_READ_ARGS(argc, argv); |
| 2858 |
READ_ARG_ULONG("-bp", mapper->bportno);
|
| 2859 |
READ_ARG_ULONG("-mbp", mapper->mbportno);
|
| 2860 |
END_READ_ARGS(); |
| 2861 |
if (mapper->bportno == -1){ |
| 2862 |
XSEGLOG2(&lc, E, "Portno for blocker must be provided");
|
| 2863 |
usage(argv[0]);
|
| 2864 |
return -1; |
| 2865 |
} |
| 2866 |
if (mapper->mbportno == -1){ |
| 2867 |
XSEGLOG2(&lc, E, "Portno for mblocker must be provided");
|
| 2868 |
usage(argv[0]);
|
| 2869 |
return -1; |
| 2870 |
} |
| 2871 |
|
| 2872 |
const struct sched_param param = { .sched_priority = 99 }; |
| 2873 |
sched_setscheduler(syscall(SYS_gettid), SCHED_FIFO, ¶m); |
| 2874 |
/* FIXME maybe place it in peer
|
| 2875 |
* should be done for each port (sportno to eportno)
|
| 2876 |
*/
|
| 2877 |
xseg_set_max_requests(peer->xseg, peer->portno_start, 5000);
|
| 2878 |
xseg_set_freequeue_size(peer->xseg, peer->portno_start, 3000, 0); |
| 2879 |
|
| 2880 |
|
| 2881 |
// test_map(peer);
|
| 2882 |
|
| 2883 |
return 0; |
| 2884 |
} |
| 2885 |
|
| 2886 |
/* FIXME this should not be here */
|
| 2887 |
int wait_reply(struct peerd *peer, struct xseg_request *expected_req) |
| 2888 |
{
|
| 2889 |
struct xseg *xseg = peer->xseg;
|
| 2890 |
xport portno_start = peer->portno_start; |
| 2891 |
xport portno_end = peer->portno_end; |
| 2892 |
struct peer_req *pr;
|
| 2893 |
xport i; |
| 2894 |
int r, c = 0; |
| 2895 |
struct xseg_request *received;
|
| 2896 |
xseg_prepare_wait(xseg, portno_start); |
| 2897 |
while(1) { |
| 2898 |
XSEGLOG2(&lc, D, "Attempting to check for reply");
|
| 2899 |
c = 1;
|
| 2900 |
while (c){
|
| 2901 |
c = 0;
|
| 2902 |
for (i = portno_start; i <= portno_end; i++) {
|
| 2903 |
received = xseg_receive(xseg, i, 0);
|
| 2904 |
if (received) {
|
| 2905 |
c = 1;
|
| 2906 |
r = xseg_get_req_data(xseg, received, (void **) &pr);
|
| 2907 |
if (r < 0 || !pr || received != expected_req){ |
| 2908 |
XSEGLOG2(&lc, W, "Received request with no pr data\n");
|
| 2909 |
xport p = xseg_respond(peer->xseg, received, peer->portno_start, X_ALLOC); |
| 2910 |
if (p == NoPort){
|
| 2911 |
XSEGLOG2(&lc, W, "Could not respond stale request");
|
| 2912 |
xseg_put_request(xseg, received, portno_start); |
| 2913 |
continue;
|
| 2914 |
} else {
|
| 2915 |
xseg_signal(xseg, p); |
| 2916 |
} |
| 2917 |
} else {
|
| 2918 |
xseg_cancel_wait(xseg, portno_start); |
| 2919 |
return 0; |
| 2920 |
} |
| 2921 |
} |
| 2922 |
} |
| 2923 |
} |
| 2924 |
xseg_wait_signal(xseg, 1000000UL);
|
| 2925 |
} |
| 2926 |
} |
| 2927 |
|
| 2928 |
|
| 2929 |
void custom_peer_finalize(struct peerd *peer) |
| 2930 |
{
|
| 2931 |
struct mapperd *mapper = __get_mapperd(peer);
|
| 2932 |
struct peer_req *pr = alloc_peer_req(peer);
|
| 2933 |
if (!pr){
|
| 2934 |
XSEGLOG2(&lc, E, "Cannot get peer request");
|
| 2935 |
return;
|
| 2936 |
} |
| 2937 |
struct map *map;
|
| 2938 |
struct xseg_request *req;
|
| 2939 |
xhash_iter_t it; |
| 2940 |
xhashidx key, val; |
| 2941 |
xhash_iter_init(mapper->hashmaps, &it); |
| 2942 |
while (xhash_iterate(mapper->hashmaps, &it, &key, &val)){
|
| 2943 |
map = (struct map *)val;
|
| 2944 |
if (!(map->flags & MF_MAP_EXCLUSIVE))
|
| 2945 |
continue;
|
| 2946 |
req = __close_map(pr, map); |
| 2947 |
if (!req)
|
| 2948 |
continue;
|
| 2949 |
wait_reply(peer, req); |
| 2950 |
if (!(req->state & XS_SERVED))
|
| 2951 |
XSEGLOG2(&lc, E, "Couldn't close map %s", map->volume);
|
| 2952 |
map->flags &= ~MF_MAP_CLOSING; |
| 2953 |
xseg_put_request(peer->xseg, req, pr->portno); |
| 2954 |
} |
| 2955 |
return;
|
| 2956 |
|
| 2957 |
|
| 2958 |
} |
| 2959 |
|
| 2960 |
void print_obj(struct map_node *mn) |
| 2961 |
{
|
| 2962 |
fprintf(stderr, "[%llu]object name: %s[%u] exists: %c\n",
|
| 2963 |
(unsigned long long) mn->objectidx, mn->object, |
| 2964 |
(unsigned int) mn->objectlen, |
| 2965 |
(mn->flags & MF_OBJECT_EXIST) ? 'y' : 'n'); |
| 2966 |
} |
| 2967 |
|
| 2968 |
void print_map(struct map *m) |
| 2969 |
{
|
| 2970 |
uint64_t nr_objs = m->size/block_size; |
| 2971 |
if (m->size % block_size)
|
| 2972 |
nr_objs++; |
| 2973 |
fprintf(stderr, "Volume name: %s[%u], size: %llu, nr_objs: %llu, version: %u\n",
|
| 2974 |
m->volume, m->volumelen, |
| 2975 |
(unsigned long long) m->size, |
| 2976 |
(unsigned long long) nr_objs, |
| 2977 |
m->version); |
| 2978 |
uint64_t i; |
| 2979 |
struct map_node *mn;
|
| 2980 |
if (nr_objs > 1000000) //FIXME to protect against invalid volume size |
| 2981 |
return;
|
| 2982 |
for (i = 0; i < nr_objs; i++) { |
| 2983 |
mn = find_object(m, i); |
| 2984 |
if (!mn){
|
| 2985 |
printf("object idx [%llu] not found!\n", (unsigned long long) i); |
| 2986 |
continue;
|
| 2987 |
} |
| 2988 |
print_obj(mn); |
| 2989 |
} |
| 2990 |
} |
| 2991 |
|
| 2992 |
/*
|
| 2993 |
void test_map(struct peerd *peer)
|
| 2994 |
{
|
| 2995 |
int i,j, ret;
|
| 2996 |
//struct sha256_ctx sha256ctx;
|
| 2997 |
unsigned char buf[SHA256_DIGEST_SIZE];
|
| 2998 |
char buf_new[XSEG_MAX_TARGETLEN + 20];
|
| 2999 |
struct map *m = malloc(sizeof(struct map));
|
| 3000 |
strncpy(m->volume, "012345678901234567890123456789ab012345678901234567890123456789ab", XSEG_MAX_TARGETLEN + 1);
|
| 3001 |
m->volume[XSEG_MAX_TARGETLEN] = 0;
|
| 3002 |
strncpy(buf_new, m->volume, XSEG_MAX_TARGETLEN);
|
| 3003 |
buf_new[XSEG_MAX_TARGETLEN + 19] = 0;
|
| 3004 |
m->volumelen = XSEG_MAX_TARGETLEN;
|
| 3005 |
m->size = 100*block_size;
|
| 3006 |
m->objects = xhash_new(3, INTEGER);
|
| 3007 |
struct map_node *map_node = calloc(100, sizeof(struct map_node));
|
| 3008 |
for (i = 0; i < 100; i++) {
|
| 3009 |
sprintf(buf_new +XSEG_MAX_TARGETLEN, "%u", i);
|
| 3010 |
gcry_md_hash_buffer(GCRY_MD_SHA256, buf, buf_new, strlen(buf_new));
|
| 3011 |
|
| 3012 |
for (j = 0; j < SHA256_DIGEST_SIZE; j++) {
|
| 3013 |
sprintf(map_node[i].object + 2*j, "%02x", buf[j]);
|
| 3014 |
}
|
| 3015 |
map_node[i].objectidx = i;
|
| 3016 |
map_node[i].objectlen = XSEG_MAX_TARGETLEN;
|
| 3017 |
map_node[i].flags = MF_OBJECT_EXIST;
|
| 3018 |
ret = insert_object(m, &map_node[i]);
|
| 3019 |
}
|
| 3020 |
|
| 3021 |
char *data = malloc(block_size);
|
| 3022 |
mapheader_to_map(m, data);
|
| 3023 |
uint64_t pos = mapheader_size;
|
| 3024 |
|
| 3025 |
for (i = 0; i < 100; i++) {
|
| 3026 |
map_node = find_object(m, i);
|
| 3027 |
if (!map_node){
|
| 3028 |
printf("no object node %d \n", i);
|
| 3029 |
exit(1);
|
| 3030 |
}
|
| 3031 |
object_to_map(data+pos, map_node);
|
| 3032 |
pos += objectsize_in_map;
|
| 3033 |
}
|
| 3034 |
// print_map(m);
|
| 3035 |
|
| 3036 |
struct map *m2 = malloc(sizeof(struct map));
|
| 3037 |
strncpy(m2->volume, "012345678901234567890123456789ab012345678901234567890123456789ab", XSEG_MAX_TARGETLEN +1);
|
| 3038 |
m->volume[XSEG_MAX_TARGETLEN] = 0;
|
| 3039 |
m->volumelen = XSEG_MAX_TARGETLEN;
|
| 3040 |
|
| 3041 |
m2->objects = xhash_new(3, INTEGER);
|
| 3042 |
ret = read_map(peer, m2, data);
|
| 3043 |
// print_map(m2);
|
| 3044 |
|
| 3045 |
int fd = open(m->volume, O_CREAT|O_WRONLY, S_IRUSR|S_IWUSR|S_IRGRP|S_IWGRP|S_IROTH);
|
| 3046 |
ssize_t r, sum = 0;
|
| 3047 |
while (sum < block_size) {
|
| 3048 |
r = write(fd, data + sum, block_size -sum);
|
| 3049 |
if (r < 0){
|
| 3050 |
perror("write");
|
| 3051 |
printf("write error\n");
|
| 3052 |
exit(1);
|
| 3053 |
}
|
| 3054 |
sum += r;
|
| 3055 |
}
|
| 3056 |
close(fd);
|
| 3057 |
map_node = find_object(m, 0);
|
| 3058 |
free(map_node);
|
| 3059 |
free(m);
|
| 3060 |
}
|
| 3061 |
*/
|