From: kernel test robot <lkp@intel.com>
To: Christoph Hellwig <hch@lst.de>
Cc: oe-kbuild-all@lists.linux.dev
Subject: [hch-misc:cap-analysis 13/13] fs/xfs/xfs_inode.c:156:6: warning: rw_semaphore 'VFS_I().i_rwsem' is not held on every path through here
Date: Sat, 20 Sep 2025 02:56:24 +0800 [thread overview]
Message-ID: <202509200215.D0fhysMM-lkp@intel.com> (raw)
tree: git://git.infradead.org/users/hch/misc.git cap-analysis
head: f0fefaa266b6cdc474b78bb5842ee7959505e166
commit: f0fefaa266b6cdc474b78bb5842ee7959505e166 [13/13] WIP: xfs: enable capability analysis
config: riscv-randconfig-002-20250920 (https://download.01.org/0day-ci/archive/20250920/202509200215.D0fhysMM-lkp@intel.com/config)
compiler: clang version 22.0.0git (https://github.com/llvm/llvm-project 7c861bcedf61607b6c087380ac711eb7ff918ca6)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250920/202509200215.D0fhysMM-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202509200215.D0fhysMM-lkp@intel.com/
All warnings (new ones prefixed by >>):
>> fs/xfs/xfs_trans_ail.c:325:2: warning: calling function 'xfs_ail_check' requires holding spinlock 'ailp->ail_lock' exclusively [-Wthread-safety-analysis]
325 | xfs_ail_check(ailp, lip);
| ^
1 warning generated.
--
>> fs/xfs/xfs_extent_busy.c:573:26: warning: expecting spinlock 'xfs_group_hold(busyp->group)..xg_busy_extents->eb_lock' to be held at start of each loop [-Wthread-safety-analysis]
573 | struct xfs_group *xg = xfs_group_hold(busyp->group);
| ^
fs/xfs/xfs_extent_busy.c:577:3: note: spinlock acquired here
577 | spin_lock(&eb->eb_lock);
| ^
>> fs/xfs/xfs_extent_busy.c:590:3: warning: releasing spinlock 'xfs_group_hold((busyp->list.next - #undefined).group)..xg_busy_extents->eb_lock' that was not held [-Wthread-safety-analysis]
590 | spin_unlock(&eb->eb_lock);
| ^
>> fs/xfs/xfs_extent_busy.c:593:1: warning: spinlock 'xfs_group_hold(busyp->group)..xg_busy_extents->eb_lock' is not held on every path through here [-Wthread-safety-analysis]
593 | }
| ^
fs/xfs/xfs_extent_busy.c:577:3: note: spinlock acquired here
577 | spin_lock(&eb->eb_lock);
| ^
3 warnings generated.
--
>> fs/xfs/xfs_log.c:481:7: warning: expecting spinlock 'log->l_icloglock' to be held at start of each loop [-Wthread-safety-analysis]
481 | if (atomic_read(&iclog->ic_refcnt)) {
| ^
fs/xfs/xfs_log.c:490:3: note: spinlock acquired here
490 | spin_lock(&log->l_icloglock);
| ^
>> fs/xfs/xfs_log.c:486:3: warning: releasing spinlock 'log->l_icloglock' that was not held [-Wthread-safety-analysis]
486 | spin_unlock(&log->l_icloglock);
| ^
fs/xfs/xfs_log.c:493:20: warning: expecting spinlock 'log->l_icloglock' to be held at start of each loop [-Wthread-safety-analysis]
493 | } while ((iclog = iclog->ic_next) != log->l_iclog);
| ^
fs/xfs/xfs_log.c:490:3: note: spinlock acquired here
490 | spin_lock(&log->l_icloglock);
| ^
>> fs/xfs/xfs_log.c:496:1: warning: spinlock 'log->l_icloglock' is still held at the end of function [-Wthread-safety-analysis]
496 | }
| ^
fs/xfs/xfs_log.c:490:3: note: spinlock acquired here
490 | spin_lock(&log->l_icloglock);
| ^
4 warnings generated.
--
>> fs/xfs/xfs_inode.c:156:6: warning: rw_semaphore 'VFS_I().i_rwsem' is not held on every path through here [-Wthread-safety-analysis]
156 | if (lock_flags & XFS_MMAPLOCK_EXCL) {
| ^
fs/xfs/xfs_inode.c:152:3: note: rw_semaphore acquired here
152 | down_read_nested(&VFS_I(ip)->i_rwsem,
| ^
>> fs/xfs/xfs_inode.c:156:6: warning: rw_semaphore 'VFS_I().i_rwsem' is not held on every path through here [-Wthread-safety-analysis]
156 | if (lock_flags & XFS_MMAPLOCK_EXCL) {
| ^
fs/xfs/xfs_inode.c:149:3: note: rw_semaphore acquired here
149 | down_write_nested(&VFS_I(ip)->i_rwsem,
| ^
>> fs/xfs/xfs_inode.c:164:6: warning: rw_semaphore 'VFS_I().i_mapping->invalidate_lock' is not held on every path through here [-Wthread-safety-analysis]
164 | if (lock_flags & XFS_ILOCK_EXCL)
| ^
fs/xfs/xfs_inode.c:160:3: note: rw_semaphore acquired here
160 | down_read_nested(&VFS_I(ip)->i_mapping->invalidate_lock,
| ^
>> fs/xfs/xfs_inode.c:164:6: warning: rw_semaphore 'VFS_I().i_mapping->invalidate_lock' is not held on every path through here [-Wthread-safety-analysis]
164 | if (lock_flags & XFS_ILOCK_EXCL)
| ^
fs/xfs/xfs_inode.c:157:3: note: rw_semaphore acquired here
157 | down_write_nested(&VFS_I(ip)->i_mapping->invalidate_lock,
| ^
>> fs/xfs/xfs_inode.c:168:1: warning: rw_semaphore '->i_lock' is not held on every path through here [-Wthread-safety-analysis]
168 | }
| ^
fs/xfs/xfs_inode.c:167:3: note: rw_semaphore acquired here
167 | down_read_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags));
| ^
>> fs/xfs/xfs_inode.c:168:1: warning: rw_semaphore '->i_lock' is not held on every path through here [-Wthread-safety-analysis]
168 | }
| ^
fs/xfs/xfs_inode.c:165:3: note: rw_semaphore acquired here
165 | down_write_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags));
| ^
fs/xfs/xfs_inode.c:199:6: warning: rw_semaphore 'VFS_I().i_rwsem' is not held on every path through here [-Wthread-safety-analysis]
199 | if (lock_flags & XFS_MMAPLOCK_EXCL) {
| ^
fs/xfs/xfs_inode.c:195:8: note: rw_semaphore acquired here
195 | if (!down_read_trylock(&VFS_I(ip)->i_rwsem))
| ^
fs/xfs/xfs_inode.c:199:6: warning: rw_semaphore 'VFS_I().i_rwsem' is not held on every path through here [-Wthread-safety-analysis]
199 | if (lock_flags & XFS_MMAPLOCK_EXCL) {
| ^
fs/xfs/xfs_inode.c:192:8: note: rw_semaphore acquired here
192 | if (!down_write_trylock(&VFS_I(ip)->i_rwsem))
| ^
fs/xfs/xfs_inode.c:207:6: warning: rw_semaphore 'VFS_I().i_mapping->invalidate_lock' is not held on every path through here [-Wthread-safety-analysis]
207 | if (lock_flags & XFS_ILOCK_EXCL) {
| ^
fs/xfs/xfs_inode.c:203:8: note: rw_semaphore acquired here
203 | if (!down_read_trylock(&VFS_I(ip)->i_mapping->invalidate_lock))
| ^
fs/xfs/xfs_inode.c:207:6: warning: rw_semaphore 'VFS_I().i_mapping->invalidate_lock' is not held on every path through here [-Wthread-safety-analysis]
207 | if (lock_flags & XFS_ILOCK_EXCL) {
| ^
fs/xfs/xfs_inode.c:200:8: note: rw_semaphore acquired here
200 | if (!down_write_trylock(&VFS_I(ip)->i_mapping->invalidate_lock))
| ^
fs/xfs/xfs_inode.c:214:9: warning: rw_semaphore '->i_lock' is not held on every path through here [-Wthread-safety-analysis]
214 | return 1;
| ^
fs/xfs/xfs_inode.c:211:8: note: rw_semaphore acquired here
211 | if (!down_read_trylock(&ip->i_lock))
| ^
fs/xfs/xfs_inode.c:214:9: warning: rw_semaphore '->i_lock' is not held on every path through here [-Wthread-safety-analysis]
214 | return 1;
| ^
fs/xfs/xfs_inode.c:208:8: note: rw_semaphore acquired here
208 | if (!down_write_trylock(&ip->i_lock))
| ^
>> fs/xfs/xfs_inode.c:218:3: warning: releasing rw_semaphore 'VFS_I().i_mapping->invalidate_lock' that was not held [-Wthread-safety-analysis]
218 | up_write(&VFS_I(ip)->i_mapping->invalidate_lock);
| ^
fs/xfs/xfs_inode.c:220:3: warning: releasing rw_semaphore 'VFS_I().i_mapping->invalidate_lock' that was not held [-Wthread-safety-analysis]
220 | up_read(&VFS_I(ip)->i_mapping->invalidate_lock);
| ^
>> fs/xfs/xfs_inode.c:223:3: warning: releasing rw_semaphore 'VFS_I().i_rwsem' that was not held [-Wthread-safety-analysis]
223 | up_write(&VFS_I(ip)->i_rwsem);
| ^
fs/xfs/xfs_inode.c:225:3: warning: releasing rw_semaphore 'VFS_I().i_rwsem' that was not held [-Wthread-safety-analysis]
225 | up_read(&VFS_I(ip)->i_rwsem);
| ^
fs/xfs/xfs_inode.c:250:3: warning: releasing rw_semaphore 'VFS_I().i_rwsem' that was not held [-Wthread-safety-analysis]
250 | up_write(&VFS_I(ip)->i_rwsem);
| ^
fs/xfs/xfs_inode.c:252:3: warning: releasing rw_semaphore 'VFS_I().i_rwsem' that was not held [-Wthread-safety-analysis]
252 | up_read(&VFS_I(ip)->i_rwsem);
| ^
fs/xfs/xfs_inode.c:255:3: warning: releasing rw_semaphore 'VFS_I().i_mapping->invalidate_lock' that was not held [-Wthread-safety-analysis]
255 | up_write(&VFS_I(ip)->i_mapping->invalidate_lock);
| ^
fs/xfs/xfs_inode.c:257:3: warning: releasing rw_semaphore 'VFS_I().i_mapping->invalidate_lock' that was not held [-Wthread-safety-analysis]
257 | up_read(&VFS_I(ip)->i_mapping->invalidate_lock);
| ^
>> fs/xfs/xfs_inode.c:260:3: warning: releasing rw_semaphore '->i_lock' that was not held [-Wthread-safety-analysis]
260 | up_write(&ip->i_lock);
| ^
fs/xfs/xfs_inode.c:262:3: warning: releasing rw_semaphore '->i_lock' that was not held [-Wthread-safety-analysis]
262 | up_read(&ip->i_lock);
| ^
fs/xfs/xfs_inode.c:281:3: warning: releasing rw_semaphore '->i_lock' that was not held [-Wthread-safety-analysis]
281 | downgrade_write(&ip->i_lock);
| ^
fs/xfs/xfs_inode.c:282:6: warning: rw_semaphore '->i_lock' is not held on every path through here [-Wthread-safety-analysis]
282 | if (lock_flags & XFS_MMAPLOCK_EXCL)
| ^
fs/xfs/xfs_inode.c:281:3: note: rw_semaphore acquired here
281 | downgrade_write(&ip->i_lock);
| ^
fs/xfs/xfs_inode.c:283:3: warning: releasing rw_semaphore 'VFS_I().i_mapping->invalidate_lock' that was not held [-Wthread-safety-analysis]
283 | downgrade_write(&VFS_I(ip)->i_mapping->invalidate_lock);
| ^
fs/xfs/xfs_inode.c:284:6: warning: rw_semaphore 'VFS_I().i_mapping->invalidate_lock' is not held on every path through here [-Wthread-safety-analysis]
284 | if (lock_flags & XFS_IOLOCK_EXCL)
| ^
fs/xfs/xfs_inode.c:283:3: note: rw_semaphore acquired here
283 | downgrade_write(&VFS_I(ip)->i_mapping->invalidate_lock);
| ^
fs/xfs/xfs_inode.c:285:3: warning: releasing rw_semaphore 'VFS_I().i_rwsem' that was not held [-Wthread-safety-analysis]
285 | downgrade_write(&VFS_I(ip)->i_rwsem);
| ^
fs/xfs/xfs_inode.c:287:2: warning: rw_semaphore 'VFS_I().i_rwsem' is not held on every path through here [-Wthread-safety-analysis]
287 | trace_xfs_ilock_demote(ip, lock_flags, _RET_IP_);
| ^
fs/xfs/xfs_inode.c:285:3: note: rw_semaphore acquired here
285 | downgrade_write(&VFS_I(ip)->i_rwsem);
| ^
28 warnings generated.
vim +156 fs/xfs/xfs_inode.c
ca76a761ea2497 Kaixu Xia 2022-06-19 108
fa96acadf1eb71 Dave Chinner 2012-10-08 109 /*
6552321831dce8 Christoph Hellwig 2016-11-30 110 * In addition to i_rwsem in the VFS inode, the xfs inode contains 2
2433480a7e1d0c Jan Kara 2021-04-12 111 * multi-reader locks: invalidate_lock and the i_lock. This routine allows
6552321831dce8 Christoph Hellwig 2016-11-30 112 * various combinations of the locks to be obtained.
fa96acadf1eb71 Dave Chinner 2012-10-08 113 *
653c60b633a901 Dave Chinner 2015-02-23 114 * The 3 locks should always be ordered so that the IO lock is obtained first,
653c60b633a901 Dave Chinner 2015-02-23 115 * the mmap lock second and the ilock last in order to prevent deadlock.
fa96acadf1eb71 Dave Chinner 2012-10-08 116 *
653c60b633a901 Dave Chinner 2015-02-23 117 * Basic locking order:
653c60b633a901 Dave Chinner 2015-02-23 118 *
2433480a7e1d0c Jan Kara 2021-04-12 119 * i_rwsem -> invalidate_lock -> page_lock -> i_ilock
653c60b633a901 Dave Chinner 2015-02-23 120 *
c1e8d7c6a7a682 Michel Lespinasse 2020-06-08 121 * mmap_lock locking order:
653c60b633a901 Dave Chinner 2015-02-23 122 *
c1e8d7c6a7a682 Michel Lespinasse 2020-06-08 123 * i_rwsem -> page lock -> mmap_lock
2433480a7e1d0c Jan Kara 2021-04-12 124 * mmap_lock -> invalidate_lock -> page_lock
653c60b633a901 Dave Chinner 2015-02-23 125 *
c1e8d7c6a7a682 Michel Lespinasse 2020-06-08 126 * The difference in mmap_lock locking order mean that we cannot hold the
2433480a7e1d0c Jan Kara 2021-04-12 127 * invalidate_lock over syscall based read(2)/write(2) based IO. These IO paths
2433480a7e1d0c Jan Kara 2021-04-12 128 * can fault in pages during copy in/out (for buffered IO) or require the
2433480a7e1d0c Jan Kara 2021-04-12 129 * mmap_lock in get_user_pages() to map the user pages into the kernel address
2433480a7e1d0c Jan Kara 2021-04-12 130 * space for direct IO. Similarly the i_rwsem cannot be taken inside a page
2433480a7e1d0c Jan Kara 2021-04-12 131 * fault because page faults already hold the mmap_lock.
653c60b633a901 Dave Chinner 2015-02-23 132 *
653c60b633a901 Dave Chinner 2015-02-23 133 * Hence to serialise fully against both syscall and mmap based IO, we need to
2433480a7e1d0c Jan Kara 2021-04-12 134 * take both the i_rwsem and the invalidate_lock. These locks should *only* be
2433480a7e1d0c Jan Kara 2021-04-12 135 * both taken in places where we need to invalidate the page cache in a race
653c60b633a901 Dave Chinner 2015-02-23 136 * free manner (e.g. truncate, hole punch and other extent manipulation
653c60b633a901 Dave Chinner 2015-02-23 137 * functions).
fa96acadf1eb71 Dave Chinner 2012-10-08 138 */
fa96acadf1eb71 Dave Chinner 2012-10-08 139 void
fa96acadf1eb71 Dave Chinner 2012-10-08 140 xfs_ilock(
fa96acadf1eb71 Dave Chinner 2012-10-08 141 xfs_inode_t *ip,
fa96acadf1eb71 Dave Chinner 2012-10-08 142 uint lock_flags)
fa96acadf1eb71 Dave Chinner 2012-10-08 143 {
fa96acadf1eb71 Dave Chinner 2012-10-08 144 trace_xfs_ilock(ip, lock_flags, _RET_IP_);
fa96acadf1eb71 Dave Chinner 2012-10-08 145
ca76a761ea2497 Kaixu Xia 2022-06-19 146 xfs_lock_flags_assert(lock_flags);
fa96acadf1eb71 Dave Chinner 2012-10-08 147
6552321831dce8 Christoph Hellwig 2016-11-30 148 if (lock_flags & XFS_IOLOCK_EXCL) {
6552321831dce8 Christoph Hellwig 2016-11-30 149 down_write_nested(&VFS_I(ip)->i_rwsem,
6552321831dce8 Christoph Hellwig 2016-11-30 150 XFS_IOLOCK_DEP(lock_flags));
6552321831dce8 Christoph Hellwig 2016-11-30 151 } else if (lock_flags & XFS_IOLOCK_SHARED) {
6552321831dce8 Christoph Hellwig 2016-11-30 152 down_read_nested(&VFS_I(ip)->i_rwsem,
6552321831dce8 Christoph Hellwig 2016-11-30 153 XFS_IOLOCK_DEP(lock_flags));
6552321831dce8 Christoph Hellwig 2016-11-30 154 }
fa96acadf1eb71 Dave Chinner 2012-10-08 155
2433480a7e1d0c Jan Kara 2021-04-12 @156 if (lock_flags & XFS_MMAPLOCK_EXCL) {
2433480a7e1d0c Jan Kara 2021-04-12 157 down_write_nested(&VFS_I(ip)->i_mapping->invalidate_lock,
2433480a7e1d0c Jan Kara 2021-04-12 158 XFS_MMAPLOCK_DEP(lock_flags));
2433480a7e1d0c Jan Kara 2021-04-12 159 } else if (lock_flags & XFS_MMAPLOCK_SHARED) {
2433480a7e1d0c Jan Kara 2021-04-12 160 down_read_nested(&VFS_I(ip)->i_mapping->invalidate_lock,
2433480a7e1d0c Jan Kara 2021-04-12 161 XFS_MMAPLOCK_DEP(lock_flags));
2433480a7e1d0c Jan Kara 2021-04-12 162 }
653c60b633a901 Dave Chinner 2015-02-23 163
fa96acadf1eb71 Dave Chinner 2012-10-08 @164 if (lock_flags & XFS_ILOCK_EXCL)
785dd131525060 Matthew Wilcox (Oracle 2024-02-19 165) down_write_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags));
fa96acadf1eb71 Dave Chinner 2012-10-08 166 else if (lock_flags & XFS_ILOCK_SHARED)
785dd131525060 Matthew Wilcox (Oracle 2024-02-19 167) down_read_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags));
fa96acadf1eb71 Dave Chinner 2012-10-08 @168 }
fa96acadf1eb71 Dave Chinner 2012-10-08 169
fa96acadf1eb71 Dave Chinner 2012-10-08 170 /*
fa96acadf1eb71 Dave Chinner 2012-10-08 171 * This is just like xfs_ilock(), except that the caller
fa96acadf1eb71 Dave Chinner 2012-10-08 172 * is guaranteed not to sleep. It returns 1 if it gets
fa96acadf1eb71 Dave Chinner 2012-10-08 173 * the requested locks and 0 otherwise. If the IO lock is
fa96acadf1eb71 Dave Chinner 2012-10-08 174 * obtained but the inode lock cannot be, then the IO lock
fa96acadf1eb71 Dave Chinner 2012-10-08 175 * is dropped before returning.
fa96acadf1eb71 Dave Chinner 2012-10-08 176 *
fa96acadf1eb71 Dave Chinner 2012-10-08 177 * ip -- the inode being locked
fa96acadf1eb71 Dave Chinner 2012-10-08 178 * lock_flags -- this parameter indicates the inode's locks to be
fa96acadf1eb71 Dave Chinner 2012-10-08 179 * to be locked. See the comment for xfs_ilock() for a list
fa96acadf1eb71 Dave Chinner 2012-10-08 180 * of valid values.
fa96acadf1eb71 Dave Chinner 2012-10-08 181 */
fa96acadf1eb71 Dave Chinner 2012-10-08 182 int
fa96acadf1eb71 Dave Chinner 2012-10-08 183 xfs_ilock_nowait(
fa96acadf1eb71 Dave Chinner 2012-10-08 184 xfs_inode_t *ip,
fa96acadf1eb71 Dave Chinner 2012-10-08 185 uint lock_flags)
fa96acadf1eb71 Dave Chinner 2012-10-08 186 {
fa96acadf1eb71 Dave Chinner 2012-10-08 187 trace_xfs_ilock_nowait(ip, lock_flags, _RET_IP_);
fa96acadf1eb71 Dave Chinner 2012-10-08 188
ca76a761ea2497 Kaixu Xia 2022-06-19 189 xfs_lock_flags_assert(lock_flags);
fa96acadf1eb71 Dave Chinner 2012-10-08 190
fa96acadf1eb71 Dave Chinner 2012-10-08 191 if (lock_flags & XFS_IOLOCK_EXCL) {
6552321831dce8 Christoph Hellwig 2016-11-30 192 if (!down_write_trylock(&VFS_I(ip)->i_rwsem))
fa96acadf1eb71 Dave Chinner 2012-10-08 193 goto out;
fa96acadf1eb71 Dave Chinner 2012-10-08 194 } else if (lock_flags & XFS_IOLOCK_SHARED) {
6552321831dce8 Christoph Hellwig 2016-11-30 195 if (!down_read_trylock(&VFS_I(ip)->i_rwsem))
fa96acadf1eb71 Dave Chinner 2012-10-08 196 goto out;
fa96acadf1eb71 Dave Chinner 2012-10-08 197 }
653c60b633a901 Dave Chinner 2015-02-23 198
653c60b633a901 Dave Chinner 2015-02-23 199 if (lock_flags & XFS_MMAPLOCK_EXCL) {
2433480a7e1d0c Jan Kara 2021-04-12 200 if (!down_write_trylock(&VFS_I(ip)->i_mapping->invalidate_lock))
653c60b633a901 Dave Chinner 2015-02-23 201 goto out_undo_iolock;
653c60b633a901 Dave Chinner 2015-02-23 202 } else if (lock_flags & XFS_MMAPLOCK_SHARED) {
2433480a7e1d0c Jan Kara 2021-04-12 203 if (!down_read_trylock(&VFS_I(ip)->i_mapping->invalidate_lock))
653c60b633a901 Dave Chinner 2015-02-23 204 goto out_undo_iolock;
653c60b633a901 Dave Chinner 2015-02-23 205 }
653c60b633a901 Dave Chinner 2015-02-23 206
fa96acadf1eb71 Dave Chinner 2012-10-08 207 if (lock_flags & XFS_ILOCK_EXCL) {
785dd131525060 Matthew Wilcox (Oracle 2024-02-19 208) if (!down_write_trylock(&ip->i_lock))
653c60b633a901 Dave Chinner 2015-02-23 209 goto out_undo_mmaplock;
fa96acadf1eb71 Dave Chinner 2012-10-08 210 } else if (lock_flags & XFS_ILOCK_SHARED) {
785dd131525060 Matthew Wilcox (Oracle 2024-02-19 211) if (!down_read_trylock(&ip->i_lock))
653c60b633a901 Dave Chinner 2015-02-23 212 goto out_undo_mmaplock;
fa96acadf1eb71 Dave Chinner 2012-10-08 213 }
fa96acadf1eb71 Dave Chinner 2012-10-08 214 return 1;
fa96acadf1eb71 Dave Chinner 2012-10-08 215
653c60b633a901 Dave Chinner 2015-02-23 216 out_undo_mmaplock:
653c60b633a901 Dave Chinner 2015-02-23 217 if (lock_flags & XFS_MMAPLOCK_EXCL)
2433480a7e1d0c Jan Kara 2021-04-12 @218 up_write(&VFS_I(ip)->i_mapping->invalidate_lock);
653c60b633a901 Dave Chinner 2015-02-23 219 else if (lock_flags & XFS_MMAPLOCK_SHARED)
2433480a7e1d0c Jan Kara 2021-04-12 220 up_read(&VFS_I(ip)->i_mapping->invalidate_lock);
fa96acadf1eb71 Dave Chinner 2012-10-08 221 out_undo_iolock:
fa96acadf1eb71 Dave Chinner 2012-10-08 222 if (lock_flags & XFS_IOLOCK_EXCL)
6552321831dce8 Christoph Hellwig 2016-11-30 @223 up_write(&VFS_I(ip)->i_rwsem);
fa96acadf1eb71 Dave Chinner 2012-10-08 224 else if (lock_flags & XFS_IOLOCK_SHARED)
6552321831dce8 Christoph Hellwig 2016-11-30 225 up_read(&VFS_I(ip)->i_rwsem);
fa96acadf1eb71 Dave Chinner 2012-10-08 226 out:
fa96acadf1eb71 Dave Chinner 2012-10-08 227 return 0;
fa96acadf1eb71 Dave Chinner 2012-10-08 228 }
fa96acadf1eb71 Dave Chinner 2012-10-08 229
fa96acadf1eb71 Dave Chinner 2012-10-08 230 /*
fa96acadf1eb71 Dave Chinner 2012-10-08 231 * xfs_iunlock() is used to drop the inode locks acquired with
fa96acadf1eb71 Dave Chinner 2012-10-08 232 * xfs_ilock() and xfs_ilock_nowait(). The caller must pass
fa96acadf1eb71 Dave Chinner 2012-10-08 233 * in the flags given to xfs_ilock() or xfs_ilock_nowait() so
fa96acadf1eb71 Dave Chinner 2012-10-08 234 * that we know which locks to drop.
fa96acadf1eb71 Dave Chinner 2012-10-08 235 *
fa96acadf1eb71 Dave Chinner 2012-10-08 236 * ip -- the inode being unlocked
fa96acadf1eb71 Dave Chinner 2012-10-08 237 * lock_flags -- this parameter indicates the inode's locks to be
fa96acadf1eb71 Dave Chinner 2012-10-08 238 * to be unlocked. See the comment for xfs_ilock() for a list
fa96acadf1eb71 Dave Chinner 2012-10-08 239 * of valid values for this parameter.
fa96acadf1eb71 Dave Chinner 2012-10-08 240 *
fa96acadf1eb71 Dave Chinner 2012-10-08 241 */
fa96acadf1eb71 Dave Chinner 2012-10-08 242 void
fa96acadf1eb71 Dave Chinner 2012-10-08 243 xfs_iunlock(
fa96acadf1eb71 Dave Chinner 2012-10-08 244 xfs_inode_t *ip,
fa96acadf1eb71 Dave Chinner 2012-10-08 245 uint lock_flags)
fa96acadf1eb71 Dave Chinner 2012-10-08 246 {
ca76a761ea2497 Kaixu Xia 2022-06-19 247 xfs_lock_flags_assert(lock_flags);
fa96acadf1eb71 Dave Chinner 2012-10-08 248
fa96acadf1eb71 Dave Chinner 2012-10-08 249 if (lock_flags & XFS_IOLOCK_EXCL)
6552321831dce8 Christoph Hellwig 2016-11-30 250 up_write(&VFS_I(ip)->i_rwsem);
fa96acadf1eb71 Dave Chinner 2012-10-08 251 else if (lock_flags & XFS_IOLOCK_SHARED)
6552321831dce8 Christoph Hellwig 2016-11-30 252 up_read(&VFS_I(ip)->i_rwsem);
fa96acadf1eb71 Dave Chinner 2012-10-08 253
653c60b633a901 Dave Chinner 2015-02-23 254 if (lock_flags & XFS_MMAPLOCK_EXCL)
2433480a7e1d0c Jan Kara 2021-04-12 255 up_write(&VFS_I(ip)->i_mapping->invalidate_lock);
653c60b633a901 Dave Chinner 2015-02-23 256 else if (lock_flags & XFS_MMAPLOCK_SHARED)
2433480a7e1d0c Jan Kara 2021-04-12 257 up_read(&VFS_I(ip)->i_mapping->invalidate_lock);
653c60b633a901 Dave Chinner 2015-02-23 258
fa96acadf1eb71 Dave Chinner 2012-10-08 259 if (lock_flags & XFS_ILOCK_EXCL)
785dd131525060 Matthew Wilcox (Oracle 2024-02-19 @260) up_write(&ip->i_lock);
fa96acadf1eb71 Dave Chinner 2012-10-08 261 else if (lock_flags & XFS_ILOCK_SHARED)
785dd131525060 Matthew Wilcox (Oracle 2024-02-19 262) up_read(&ip->i_lock);
fa96acadf1eb71 Dave Chinner 2012-10-08 263
fa96acadf1eb71 Dave Chinner 2012-10-08 264 trace_xfs_iunlock(ip, lock_flags, _RET_IP_);
fa96acadf1eb71 Dave Chinner 2012-10-08 265 }
fa96acadf1eb71 Dave Chinner 2012-10-08 266
:::::: The code at line 156 was first introduced by commit
:::::: 2433480a7e1d0c057442b284c336cfaa61523117 xfs: Convert to use invalidate_lock
:::::: TO: Jan Kara <jack@suse.cz>
:::::: CC: Jan Kara <jack@suse.cz>
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
reply other threads:[~2025-09-19 18:57 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=202509200215.D0fhysMM-lkp@intel.com \
--to=lkp@intel.com \
--cc=hch@lst.de \
--cc=oe-kbuild-all@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.