[kdave-btrfs-devel:misc-next 68/68] fs/btrfs/ioctl.c:1192: warning: This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
7:05 a.m.
tree: https://github.com/kdave/btrfs-devel.git misc-next
head: e6c69fcbee7ef1d7bde4ff78eb1377dbe09d71cf
commit: e6c69fcbee7ef1d7bde4ff78eb1377dbe09d71cf [68/68] btrfs: defrag: use control
structure in btrfs_defrag_file()
config: arm64-randconfig-r036-20220214
(https://download.01.org/0day-ci/archive/20220216/202202160409.lvaZWYmg-lk...)
compiler: clang version 15.0.0 (https://github.com/llvm/llvm-project
37f422f4ac31c8b8041c6b62065263314282dab6)
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O
~/bin/make.cross
chmod +x ~/bin/make.cross
# install arm64 cross compiling tool for clang build
# apt-get install binutils-aarch64-linux-gnu
#
https://github.com/kdave/btrfs-devel/commit/e6c69fcbee7ef1d7bde4ff78eb137...
git remote add kdave-btrfs-devel https://github.com/kdave/btrfs-devel.git
git fetch --no-tags kdave-btrfs-devel misc-next
git checkout e6c69fcbee7ef1d7bde4ff78eb1377dbe09d71cf
# save the config file to linux build tree
mkdir build_dir
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir
ARCH=arm64 SHELL=/bin/bash drivers/mfd/ fs/btrfs/ mm/kasan/
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp(a)intel.com>
All warnings (new ones prefixed by >>):
> fs/btrfs/ioctl.c:1192: warning: This comment starts with
'/**', but isn't a kernel-doc comment. Refer
Documentation/doc-guide/kernel-doc.rst
* Collect all valid target extents.
fs/btrfs/ioctl.c:1598: warning: This comment starts with '/**', but isn't a
kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst
* Entry point to file defragmentation.
vim +1192 fs/btrfs/ioctl.c
4cb5300bc839b8a Chris Mason 2011-05-24 1190
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1191 /**
eb793cf857828db Qu Wenruo 2021-08-06 @1192 * Collect all valid target extents.
eb793cf857828db Qu Wenruo 2021-08-06 1193 *
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1194 * @inode: inode to defragment
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1195 * @ctrl: extra defrag policy control
eb793cf857828db Qu Wenruo 2021-08-06 1196 * @start: file offset to lookup
eb793cf857828db Qu Wenruo 2021-08-06 1197 * @len: length to lookup
e9eec72151e215c Qu Wenruo 2021-08-06 1198 * @locked: if the range has already
held extent lock
eb793cf857828db Qu Wenruo 2021-08-06 1199 * @target_list: list of targets file
extents
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1200 *
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1201 * Will update ctrl::last_scanned.
4cb5300bc839b8a Chris Mason 2011-05-24 1202 */
eb793cf857828db Qu Wenruo 2021-08-06 1203 static int defrag_collect_targets(struct
btrfs_inode *inode,
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1204 struct btrfs_defrag_ctrl *ctrl,
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1205 u64 start, u32 len,
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1206 bool locked, struct list_head
*target_list)
eb793cf857828db Qu Wenruo 2021-08-06 1207 {
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1208 bool do_compress = (ctrl->flags &
BTRFS_DEFRAG_RANGE_COMPRESS);
c0a67d1ead1393b Qu Wenruo 2022-02-11 1209 bool last_is_target = false;
eb793cf857828db Qu Wenruo 2021-08-06 1210 u64 cur = start;
eb793cf857828db Qu Wenruo 2021-08-06 1211 int ret = 0;
4cb5300bc839b8a Chris Mason 2011-05-24 1212
eb793cf857828db Qu Wenruo 2021-08-06 1213 while (cur < start + len) {
eb793cf857828db Qu Wenruo 2021-08-06 1214 struct extent_map *em;
eb793cf857828db Qu Wenruo 2021-08-06 1215 struct defrag_target_range *new;
eb793cf857828db Qu Wenruo 2021-08-06 1216 bool next_mergeable = true;
eb793cf857828db Qu Wenruo 2021-08-06 1217 u64 range_len;
eb793cf857828db Qu Wenruo 2021-08-06 1218
c0a67d1ead1393b Qu Wenruo 2022-02-11 1219 last_is_target = false;
e9eec72151e215c Qu Wenruo 2021-08-06 1220 em =
defrag_lookup_extent(&inode->vfs_inode, cur, locked);
eb793cf857828db Qu Wenruo 2021-08-06 1221 if (!em)
eb793cf857828db Qu Wenruo 2021-08-06 1222 break;
eb793cf857828db Qu Wenruo 2021-08-06 1223
eb793cf857828db Qu Wenruo 2021-08-06 1224 /* Skip hole/inline/preallocated extents
*/
eb793cf857828db Qu Wenruo 2021-08-06 1225 if (em->block_start >=
EXTENT_MAP_LAST_BYTE ||
eb793cf857828db Qu Wenruo 2021-08-06 1226 test_bit(EXTENT_FLAG_PREALLOC,
&em->flags))
eb793cf857828db Qu Wenruo 2021-08-06 1227 goto next;
4cb5300bc839b8a Chris Mason 2011-05-24 1228
eb793cf857828db Qu Wenruo 2021-08-06 1229 /* Skip older extent */
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1230 if (em->generation <
ctrl->newer_than)
eb793cf857828db Qu Wenruo 2021-08-06 1231 goto next;
7f458a3873ae94e Filipe Manana 2020-11-04 1232
79c19b4d71f61a3 Qu Wenruo 2022-02-08 1233 /* This em is under writeback, no need
to defrag */
79c19b4d71f61a3 Qu Wenruo 2022-02-08 1234 if (em->generation == (u64)-1)
79c19b4d71f61a3 Qu Wenruo 2022-02-08 1235 goto next;
79c19b4d71f61a3 Qu Wenruo 2022-02-08 1236
0cb5950f3f3b51a Filipe Manana 2022-01-20 1237 /*
0cb5950f3f3b51a Filipe Manana 2022-01-20 1238 * Our start offset might be in the
middle of an existing extent
0cb5950f3f3b51a Filipe Manana 2022-01-20 1239 * map, so take that into account.
0cb5950f3f3b51a Filipe Manana 2022-01-20 1240 */
0cb5950f3f3b51a Filipe Manana 2022-01-20 1241 range_len = em->len - (cur -
em->start);
0cb5950f3f3b51a Filipe Manana 2022-01-20 1242 /*
0cb5950f3f3b51a Filipe Manana 2022-01-20 1243 * If this range of the extent map is
already flagged for delalloc,
0cb5950f3f3b51a Filipe Manana 2022-01-20 1244 * skip it, because:
0cb5950f3f3b51a Filipe Manana 2022-01-20 1245 *
0cb5950f3f3b51a Filipe Manana 2022-01-20 1246 * 1) We could deadlock later, when
trying to reserve space for
0cb5950f3f3b51a Filipe Manana 2022-01-20 1247 * delalloc, because in case we
can't immediately reserve space
0cb5950f3f3b51a Filipe Manana 2022-01-20 1248 * the flusher can start delalloc and
wait for the respective
0cb5950f3f3b51a Filipe Manana 2022-01-20 1249 * ordered extents to complete. The
deadlock would happen
0cb5950f3f3b51a Filipe Manana 2022-01-20 1250 * because we do the space
reservation while holding the range
0cb5950f3f3b51a Filipe Manana 2022-01-20 1251 * locked, and starting writeback, or
finishing an ordered
0cb5950f3f3b51a Filipe Manana 2022-01-20 1252 * extent, requires locking the
range;
0cb5950f3f3b51a Filipe Manana 2022-01-20 1253 *
0cb5950f3f3b51a Filipe Manana 2022-01-20 1254 * 2) If there's delalloc there, it
means there's dirty pages for
0cb5950f3f3b51a Filipe Manana 2022-01-20 1255 * which writeback has not started
yet (we clean the delalloc
0cb5950f3f3b51a Filipe Manana 2022-01-20 1256 * flag when starting writeback and
after creating an ordered
0cb5950f3f3b51a Filipe Manana 2022-01-20 1257 * extent). If we mark pages in an
adjacent range for defrag,
0cb5950f3f3b51a Filipe Manana 2022-01-20 1258 * then we will have a larger
contiguous range for delalloc,
0cb5950f3f3b51a Filipe Manana 2022-01-20 1259 * very likely resulting in a larger
extent after writeback is
0cb5950f3f3b51a Filipe Manana 2022-01-20 1260 * triggered (except in a case of
free space fragmentation).
0cb5950f3f3b51a Filipe Manana 2022-01-20 1261 */
0cb5950f3f3b51a Filipe Manana 2022-01-20 1262 if
(test_range_bit(&inode->io_tree, cur, cur + range_len - 1,
0cb5950f3f3b51a Filipe Manana 2022-01-20 1263 EXTENT_DELALLOC, 0, NULL))
0cb5950f3f3b51a Filipe Manana 2022-01-20 1264 goto next;
0cb5950f3f3b51a Filipe Manana 2022-01-20 1265
7f458a3873ae94e Filipe Manana 2020-11-04 1266 /*
eb793cf857828db Qu Wenruo 2021-08-06 1267 * For do_compress case, we want to
compress all valid file
eb793cf857828db Qu Wenruo 2021-08-06 1268 * extents, thus no @extent_thresh or
mergeable check.
7f458a3873ae94e Filipe Manana 2020-11-04 1269 */
eb793cf857828db Qu Wenruo 2021-08-06 1270 if (do_compress)
eb793cf857828db Qu Wenruo 2021-08-06 1271 goto add;
eb793cf857828db Qu Wenruo 2021-08-06 1272
eb793cf857828db Qu Wenruo 2021-08-06 1273 /* Skip too large extent */
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1274 if (range_len >=
ctrl->extent_thresh)
eb793cf857828db Qu Wenruo 2021-08-06 1275 goto next;
eb793cf857828db Qu Wenruo 2021-08-06 1276
c55eaee8366f0b9 Qu Wenruo 2022-01-28 1277 /*
c55eaee8366f0b9 Qu Wenruo 2022-01-28 1278 * Skip extents already at its max
capacity, this is mostly for
c55eaee8366f0b9 Qu Wenruo 2022-01-28 1279 * compressed extents, which max cap is
only 128K.
c55eaee8366f0b9 Qu Wenruo 2022-01-28 1280 */
c55eaee8366f0b9 Qu Wenruo 2022-01-28 1281 if (em->len >=
get_extent_max_capacity(em))
c55eaee8366f0b9 Qu Wenruo 2022-01-28 1282 goto next;
c55eaee8366f0b9 Qu Wenruo 2022-01-28 1283
e9eec72151e215c Qu Wenruo 2021-08-06 1284 next_mergeable =
defrag_check_next_extent(&inode->vfs_inode, em,
e9eec72151e215c Qu Wenruo 2021-08-06 1285 locked);
eb793cf857828db Qu Wenruo 2021-08-06 1286 if (!next_mergeable) {
eb793cf857828db Qu Wenruo 2021-08-06 1287 struct defrag_target_range *last;
eb793cf857828db Qu Wenruo 2021-08-06 1288
eb793cf857828db Qu Wenruo 2021-08-06 1289 /* Empty target list, no way to merge
with last entry */
eb793cf857828db Qu Wenruo 2021-08-06 1290 if (list_empty(target_list))
eb793cf857828db Qu Wenruo 2021-08-06 1291 goto next;
eb793cf857828db Qu Wenruo 2021-08-06 1292 last =
list_entry(target_list->prev,
eb793cf857828db Qu Wenruo 2021-08-06 1293 struct defrag_target_range, list);
eb793cf857828db Qu Wenruo 2021-08-06 1294 /* Not mergeable with last entry */
eb793cf857828db Qu Wenruo 2021-08-06 1295 if (last->start + last->len !=
cur)
eb793cf857828db Qu Wenruo 2021-08-06 1296 goto next;
7f458a3873ae94e Filipe Manana 2020-11-04 1297
eb793cf857828db Qu Wenruo 2021-08-06 1298 /* Mergeable, fall through to add it to
@target_list. */
7f458a3873ae94e Filipe Manana 2020-11-04 1299 }
eb793cf857828db Qu Wenruo 2021-08-06 1300
eb793cf857828db Qu Wenruo 2021-08-06 1301 add:
c0a67d1ead1393b Qu Wenruo 2022-02-11 1302 last_is_target = true;
eb793cf857828db Qu Wenruo 2021-08-06 1303 range_len = min(extent_map_end(em),
start + len) - cur;
eb793cf857828db Qu Wenruo 2021-08-06 1304 /*
eb793cf857828db Qu Wenruo 2021-08-06 1305 * This one is a good target, check if
it can be merged into
eb793cf857828db Qu Wenruo 2021-08-06 1306 * last range of the target list.
eb793cf857828db Qu Wenruo 2021-08-06 1307 */
eb793cf857828db Qu Wenruo 2021-08-06 1308 if (!list_empty(target_list)) {
eb793cf857828db Qu Wenruo 2021-08-06 1309 struct defrag_target_range *last;
eb793cf857828db Qu Wenruo 2021-08-06 1310
eb793cf857828db Qu Wenruo 2021-08-06 1311 last =
list_entry(target_list->prev,
eb793cf857828db Qu Wenruo 2021-08-06 1312 struct defrag_target_range, list);
eb793cf857828db Qu Wenruo 2021-08-06 1313 ASSERT(last->start + last->len
<= cur);
eb793cf857828db Qu Wenruo 2021-08-06 1314 if (last->start + last->len ==
cur) {
eb793cf857828db Qu Wenruo 2021-08-06 1315 /* Mergeable, enlarge the last entry
*/
eb793cf857828db Qu Wenruo 2021-08-06 1316 last->len += range_len;
eb793cf857828db Qu Wenruo 2021-08-06 1317 goto next;
7f458a3873ae94e Filipe Manana 2020-11-04 1318 }
eb793cf857828db Qu Wenruo 2021-08-06 1319 /* Fall through to allocate a new entry
*/
eb793cf857828db Qu Wenruo 2021-08-06 1320 }
eb793cf857828db Qu Wenruo 2021-08-06 1321
eb793cf857828db Qu Wenruo 2021-08-06 1322 /* Allocate new defrag_target_range */
eb793cf857828db Qu Wenruo 2021-08-06 1323 new = kmalloc(sizeof(*new), GFP_NOFS);
eb793cf857828db Qu Wenruo 2021-08-06 1324 if (!new) {
7f458a3873ae94e Filipe Manana 2020-11-04 1325 free_extent_map(em);
eb793cf857828db Qu Wenruo 2021-08-06 1326 ret = -ENOMEM;
eb793cf857828db Qu Wenruo 2021-08-06 1327 break;
7f458a3873ae94e Filipe Manana 2020-11-04 1328 }
eb793cf857828db Qu Wenruo 2021-08-06 1329 new->start = cur;
eb793cf857828db Qu Wenruo 2021-08-06 1330 new->len = range_len;
eb793cf857828db Qu Wenruo 2021-08-06 1331 list_add_tail(&new->list,
target_list);
7f458a3873ae94e Filipe Manana 2020-11-04 1332
eb793cf857828db Qu Wenruo 2021-08-06 1333 next:
eb793cf857828db Qu Wenruo 2021-08-06 1334 cur = extent_map_end(em);
eb793cf857828db Qu Wenruo 2021-08-06 1335 free_extent_map(em);
eb793cf857828db Qu Wenruo 2021-08-06 1336 }
eb793cf857828db Qu Wenruo 2021-08-06 1337 if (ret < 0) {
eb793cf857828db Qu Wenruo 2021-08-06 1338 struct defrag_target_range *entry;
eb793cf857828db Qu Wenruo 2021-08-06 1339 struct defrag_target_range *tmp;
4cb5300bc839b8a Chris Mason 2011-05-24 1340
eb793cf857828db Qu Wenruo 2021-08-06 1341 list_for_each_entry_safe(entry, tmp,
target_list, list) {
eb793cf857828db Qu Wenruo 2021-08-06 1342 list_del_init(&entry->list);
eb793cf857828db Qu Wenruo 2021-08-06 1343 kfree(entry);
eb793cf857828db Qu Wenruo 2021-08-06 1344 }
eb793cf857828db Qu Wenruo 2021-08-06 1345 }
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1346 if (!ret) {
c0a67d1ead1393b Qu Wenruo 2022-02-11 1347 /*
c0a67d1ead1393b Qu Wenruo 2022-02-11 1348 * If the last extent is not a target,
the caller can skip to
c0a67d1ead1393b Qu Wenruo 2022-02-11 1349 * the end of that extent.
c0a67d1ead1393b Qu Wenruo 2022-02-11 1350 * Otherwise, we can only go the end of
the specified range.
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1351 *
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1352 * And we may got a range smaller than
current
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1353 * ctrl->last_scanned (e.g. executed
in the defrag_one_range
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1354 * call), so we have to ensure we
didn't decrease
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1355 * ctrl->last_scanned.
c0a67d1ead1393b Qu Wenruo 2022-02-11 1356 */
c0a67d1ead1393b Qu Wenruo 2022-02-11 1357 if (!last_is_target)
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1358 ctrl->last_scanned = max(cur,
ctrl->last_scanned);
c0a67d1ead1393b Qu Wenruo 2022-02-11 1359 else
e6c69fcbee7ef1d Qu Wenruo 2022-02-11 1360 ctrl->last_scanned = max(start +
len, ctrl->last_scanned);
c0a67d1ead1393b Qu Wenruo 2022-02-11 1361 }
eb793cf857828db Qu Wenruo 2021-08-06 1362 return ret;
4cb5300bc839b8a Chris Mason 2011-05-24 1363 }
4cb5300bc839b8a Chris Mason 2011-05-24 1364
:::::: The code at line 1192 was first introduced by commit
:::::: eb793cf857828dbb1f21bfe405e5e493fdceae6c btrfs: defrag: introduce helper to collect
target file extents
:::::: TO: Qu Wenruo <wqu(a)suse.com>
:::::: CC: David Sterba <dsterba(a)suse.com>
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
217
days inactive
217
days old
0 comments
1 participants
participants (1)
-
kernel test robot