diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2015-04-16 23:27:56 -0400 |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2015-04-16 23:27:56 -0400 |
commit | 4fc8adcfec3da639da76e8314c9ccefe5bf9a045 (patch) | |
tree | e07a2dea8acf04d8bbbecd4fd3a571653ecdd953 /fs | |
parent | 84588e7a5d8220446d677d7b909a20ee7a4496b9 (diff) | |
parent | aa4d86163e4e91a1ac560954a554bab417e338f4 (diff) |
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull third hunk of vfs changes from Al Viro:
"This contains the ->direct_IO() changes from Omar + saner
generic_write_checks() + dealing with fcntl()/{read,write}() races
(mirroring O_APPEND/O_DIRECT into iocb->ki_flags and instead of
repeatedly looking at ->f_flags, which can be changed by fcntl(2),
check ->ki_flags - which cannot) + infrastructure bits for dhowells'
d_inode annotations + Christophs switch of /dev/loop to
vfs_iter_write()"
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (30 commits)
block: loop: switch to VFS ITER_BVEC
configfs: Fix inconsistent use of file_inode() vs file->f_path.dentry->d_inode
VFS: Make pathwalk use d_is_reg() rather than S_ISREG()
VFS: Fix up debugfs to use d_is_dir() in place of S_ISDIR()
VFS: Combine inode checks with d_is_negative() and d_is_positive() in pathwalk
NFS: Don't use d_inode as a variable name
VFS: Impose ordering on accesses of d_inode and d_flags
VFS: Add owner-filesystem positive/negative dentry checks
nfs: generic_write_checks() shouldn't be done on swapout...
ocfs2: use __generic_file_write_iter()
mirror O_APPEND and O_DIRECT into iocb->ki_flags
switch generic_write_checks() to iocb and iter
ocfs2: move generic_write_checks() before the alignment checks
ocfs2_file_write_iter: stop messing with ppos
udf_file_write_iter: reorder and simplify
fuse: ->direct_IO() doesn't need generic_write_checks()
ext4_file_write_iter: move generic_write_checks() up
xfs_file_aio_write_checks: switch to iocb/iov_iter
generic_write_checks(): drop isblk argument
blkdev_write_iter: expand generic_file_checks() call in there
...
Diffstat (limited to 'fs')
45 files changed, 439 insertions, 562 deletions
diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c index 2e38f9a5b472..be35d05a4d0e 100644 --- a/fs/9p/vfs_addr.c +++ b/fs/9p/vfs_addr.c @@ -230,7 +230,6 @@ static int v9fs_launder_page(struct page *page) /** * v9fs_direct_IO - 9P address space operation for direct I/O - * @rw: direction (read or write) * @iocb: target I/O control block * @iov: array of vectors that define I/O buffer * @pos: offset in file to begin the operation @@ -248,12 +247,12 @@ static int v9fs_launder_page(struct page *page) * */ static ssize_t -v9fs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, loff_t pos) +v9fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter, loff_t pos) { struct file *file = iocb->ki_filp; ssize_t n; int err = 0; - if (rw & WRITE) { + if (iov_iter_rw(iter) == WRITE) { n = p9_client_write(file->private_data, pos, iter, &err); if (n) { struct inode *inode = file_inode(file); diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index d7fcb775311e..2a9dd37dc426 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -404,21 +404,16 @@ static ssize_t v9fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct file *file = iocb->ki_filp; - ssize_t retval = 0; - loff_t origin = iocb->ki_pos; - size_t count = iov_iter_count(from); + ssize_t retval; + loff_t origin; int err = 0; - retval = generic_write_checks(file, &origin, &count, 0); - if (retval) + retval = generic_write_checks(iocb, from); + if (retval <= 0) return retval; - iov_iter_truncate(from, count); - - if (!count) - return 0; - - retval = p9_client_write(file->private_data, origin, from, &err); + origin = iocb->ki_pos; + retval = p9_client_write(file->private_data, iocb->ki_pos, from, &err); if (retval > 0) { struct inode *inode = file_inode(file); loff_t i_size; @@ -428,12 +423,11 @@ v9fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) if (inode->i_mapping && inode->i_mapping->nrpages) invalidate_inode_pages2_range(inode->i_mapping, pg_start, pg_end); - origin += retval; + iocb->ki_pos += retval; i_size = i_size_read(inode); - iocb->ki_pos = origin; - if (origin > i_size) { - inode_add_bytes(inode, origin - i_size); - i_size_write(inode, origin); + if (iocb->ki_pos > i_size) { + inode_add_bytes(inode, iocb->ki_pos - i_size); + i_size_write(inode, iocb->ki_pos); } return retval; } diff --git a/fs/affs/file.c b/fs/affs/file.c index 7c1a3d4c19c2..dcf27951781c 100644 --- a/fs/affs/file.c +++ b/fs/affs/file.c @@ -389,8 +389,7 @@ static void affs_write_failed(struct address_space *mapping, loff_t to) } static ssize_t -affs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, - loff_t offset) +affs_direct_IO(struct kiocb *iocb, struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct address_space *mapping = file->f_mapping; @@ -398,15 +397,15 @@ affs_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, size_t count = iov_iter_count(iter); ssize_t ret; - if (rw == WRITE) { + if (iov_iter_rw(iter) == WRITE) { loff_t size = offset + count; if (AFFS_I(inode)->mmu_private < size) return 0; } - ret = blockdev_direct_IO(rw, iocb, inode, iter, offset, affs_get_block); - if (ret < 0 && (rw & WRITE)) + ret = blockdev_direct_IO(iocb, inode, iter, offset, affs_get_block); + if (ret < 0 && iov_iter_rw(iter) == WRITE) affs_write_failed(mapping, offset + count); return ret; } @@ -1517,7 +1517,7 @@ static int io_submit_one(struct kioctx *ctx, struct iocb __user *user_iocb, } req->common.ki_pos = iocb->aio_offset; req->common.ki_complete = aio_complete; - req->common.ki_flags = 0; + req->common.ki_flags = iocb_flags(req->common.ki_filp); if (iocb->aio_flags & IOCB_FLAG_RESFD) { /* diff --git a/fs/block_dev.c b/fs/block_dev.c index b5e87896f517..897ee0503932 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -146,15 +146,13 @@ blkdev_get_block(struct inode *inode, sector_t iblock, } static ssize_t -blkdev_direct_IO(int rw, struct kiocb *iocb, struct iov_iter *iter, - loff_t offset) +blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter, loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; - return __blockdev_direct_IO(rw, iocb, inode, I_BDEV(inode), iter, - offset, blkdev_get_block, - NULL, NULL, 0); + return __blockdev_direct_IO(iocb, inode, I_BDEV(inode), iter, offset, + blkdev_get_block, NULL, NULL, 0); } int __sync_blockdev(struct block_device *bdev, int wait) @@ -1597,9 +1595,22 @@ static long block_ioctl(struct file *file, unsigned cmd, unsigned long arg) ssize_t blkdev_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct file *file = iocb->ki_filp; + struct inode *bd_inode = file->f_mapping->host; + loff_t size = i_size_read(bd_inode); struct blk_plug plug; ssize_t ret; + if (bdev_read_only(I_BDEV(bd_inode))) + return -EPERM; + + if (!iov_iter_count(from)) + return 0; + + if (iocb->ki_pos >= size) + return -ENOSPC; + + iov_iter_truncate(from, size - iocb->ki_pos); + blk_start_plug(&plug); ret = __generic_file_write_iter(iocb, from); if (ret > 0) { diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index cdc801c85105..faa7d390841b 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1739,27 +1739,19 @@ static ssize_t btrfs_file_write_iter(struct kiocb *iocb, u64 start_pos; u64 end_pos; ssize_t num_written = 0; - ssize_t err = 0; - size_t count = iov_iter_count(from); bool sync = (file->f_flags & O_DSYNC) || IS_SYNC(file->f_mapping->host); - loff_t pos = iocb->ki_pos; + ssize_t err; + loff_t pos; + size_t count; mutex_lock(&inode->i_mutex); - - current->backing_dev_info = inode_to_bdi(inode); - err = generic_write_checks(file, &pos, &count, S_ISBLK(inode->i_mode)); - if (err) { + err = generic_write_checks(iocb, from); + if (err <= 0) { mutex_unlock(&inode->i_mutex); - goto out; - } - - if (count == 0) { - mutex_unlock(&inode->i_mutex); - goto out; + return err; } - iov_iter_truncate(from, count); - + current->backing_dev_info = inode_to_bdi(inode); err = file_remove_suid(file); if (err) { mutex_unlock(&inode->i_mutex); @@ -1786,6 +1778,8 @@ static ssize_t btrfs_file_write_iter(struct kiocb *iocb, */ update_time_for_write(inode); + pos = iocb->ki_pos; + count = iov_iter_count(from); start_pos = round_down(pos, root->sectorsize); if (start_pos > i_size_read(inode)) { /* Expand hole size to cover write data, preventing empty gap */ @@ -1800,7 +1794,7 @@ static ssize_t btrfs_file_write_iter(struct kiocb *iocb, if (sync) atomic_inc(&BTRFS_I(inode)->sync_writers); - if (file->f_flags & O_DIRECT) { + if (iocb->ki_flags & IOCB_DIRECT) { num_written = __btrfs_direct_write(iocb, from, pos); } else { num_written = __btrfs_buffered_write(file, from, pos); diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 686331f22b15..43192e10cc43 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -8081,7 +8081,7 @@ free_ordered: bio_endio(dio_bio, ret); } -static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *iocb, +static ssize_t check_direct_IO(struct btrfs_root *root, struct kiocb *iocb, const struct iov_iter *iter, loff_t offset) { int seg; @@ -8096,7 +8096,7 @@ static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *io goto out; /* If this is a write we don't need to check anymore */ - if (rw & WRITE) + if (iov_iter_rw(iter) == WRITE) return 0; /* * Check to make sure we don't have duplicate iov_base's in this @@ -8114,8 +8114,8 @@ out: return retval; } -static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, - struct iov_iter *iter, loff_t offset) +static ssize_t btrfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter, + loff_t offset) { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; @@ -8126,7 +8126,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, bool relock = false; ssize_t ret; - if (check_direct_IO(BTRFS_I(inode)->root, rw, iocb, iter, offset)) + if (check_direct_IO(BTRFS_I(inode)->root, iocb, iter, offset)) return 0; atomic_inc(&inode->i_dio_count); @@ -8144,7 +8144,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, filemap_fdatawrite_range(inode->i_mapping, offset, offset + count - 1); - if (rw & WRITE) { + if (iov_iter_rw(iter) == WRITE) { /* * If the write DIO is beyond the EOF, we need update * the isize, but it is protected by i_mutex. So we can @@ -8174,11 +8174,11 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, wakeup = false; } - ret = __blockdev_direct_IO(rw, iocb, inode, - BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev, - iter, offset, btrfs_get_blocks_direct, NULL, - btrfs_submit_direct, flags); - if (rw & WRITE) { + ret = __blockdev_direct_IO(iocb, inode, + BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev, + iter, offset, btrfs_get_blocks_direct, NULL, + btrfs_submit_direct, flags); + if (iov_iter_rw(iter) == WRITE) { current->journal_info = NULL; if (ret < 0 && ret != -EIOCBQUEUED) btrfs_delalloc_release_space(inode, count); diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index fd5599d32362..155ab9c0246b 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -1198,8 +1198,7 @@ static int ceph_write_end(struct file *file, struct address_space *mapping, * intercept O_DIRECT reads and writes early, this function should * never get called. */ -static ssize_t ceph_direct_io(int rw, struct kiocb *iocb, - struct iov_iter *iter, +static ssize_t ceph_direct_io(struct kiocb *iocb, struct iov_iter *iter, loff_t pos) { WARN_ON(1); diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 56237ea5fc22..b9b8eb225f66 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -457,7 +457,7 @@ static ssize_t ceph_sync_read(struct kiocb *iocb, struct iov_iter *i, if (ret < 0) return ret; - if (file->f_flags & O_DIRECT) { + if (iocb->ki_flags & IOCB_DIRECT) { while (iov_iter_count(i)) { size_t start; ssize_t n; @@ -828,7 +828,7 @@ again: return ret; if ((got & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) == 0 || - (iocb->ki_filp->f_flags & O_DIRECT) || + (iocb->ki_flags & IOCB_DIRECT) || (fi->flags & CEPH_F_SYNC)) { dout("aio_sync_read %p %llx.%llx %llu~%u got cap refs on %s\n", @@ -941,9 +941,9 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from) struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_osd_client *osdc = &ceph_sb_to_client(inode->i_sb)->client->osdc; - ssize_t count = iov_iter_count(from), written = 0; + ssize_t count, written = 0; int err, want, got; - loff_t pos = iocb->ki_pos; + loff_t pos; if (ceph_snap(inode) != CEPH_NOSNAP) return -EROFS; @@ -953,14 +953,12 @@ static ssize_t ceph_write_iter(struct kiocb *iocb, struct iov_iter *from) /* We can write back this queue in page reclaim */ current->backing_dev_info = inode_to_bdi(inode); - err = generic_write_checks(file, &pos, &count, S_ISBLK(inode->i_mode)); - if (err) - goto out; - - if (count == 0) + err = generic_write_checks(iocb, from); + if (err <= 0) goto out; - iov_iter_truncate(from, count); + pos = iocb->ki_pos; + count = iov_iter_count(from); err = file_remove_suid(file); if (err) goto out; @@ -997,12 +995,12 @@ retry_snap: inode, ceph_vinop(inode), pos, count, ceph_cap_string(got)); if ((got & (CEPH_CAP_FILE_BUFFER|CEPH_CAP_FILE_LAZYIO)) == 0 || - (file->f_flags & O_DIRECT) || (fi->flags & CEPH_F_SYNC)) { + (iocb->ki_flags & IOCB_DIRECT) || (fi->flags & CEPH_F_SYNC)) { struct iov_iter data; mutex_unlock(&inode->i_mutex); /* we might need to revert back to that point */ data = *from; - if (file->f_flags & O_DIRECT) + if (iocb->ki_flags & IOCB_DIRECT) written = ceph_sync_direct_write(iocb, &data, pos); else written = ceph_sync_write(iocb, &data, pos); diff --git a/fs/cifs/file.c b/fs/cifs/file.c index ca30c391a894..ca2bc5406306 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -2560,10 +2560,9 @@ cifs_write_from_iter(loff_t offset, size_t len, struct iov_iter *from, return rc; } -static ssize_t -cifs_iovec_write(struct file *file, struct iov_iter *from, loff_t *poffset) +ssize_t cifs_user_writev(struct kiocb *iocb, struct iov_iter *from) { - size_t len; + struct file *file = iocb->ki_filp; ssize_t total_written = 0; struct cifsFileInfo *open_file; struct cifs_tcon *tcon; @@ -2573,15 +2572,15 @@ cifs_iovec_write(struct file *file, struct iov_iter *from, loff_t *poffset) struct iov_iter saved_from; int rc; - len = iov_iter_count(from); - rc = generic_write_checks(file, poffset, &len, 0); - if (rc) - return rc; - - if (!len) - return 0; + /* + * BB - optimize the way when signing is disabled. We can drop this + * extra memory-to-memory copying and use iovec buffers for constructing + * write request. + */ - iov_iter_truncate(from, len); + rc = generic_write_checks(iocb, from); + if (rc <= 0) + return rc; INIT_LIST_HEAD(&wdata_list); cifs_sb = CIFS_FILE_SB(file); @@ -2593,8 +2592,8 @@ cifs_iovec_write(struct file *file, struct iov_iter *from, loff_t *poffset) memcpy(&saved_from, from, sizeof(struct iov_iter)); - rc = cifs_write_from_iter(*poffset, len, from, open_file, cifs_sb, - &wdata_list); + rc = cifs_write_from_iter(iocb->ki_pos, iov_iter_count(from), from, + open_file, cifs_sb, &wdata_list); /* * If at least one write was successfully sent, then discard any rc @@ -2633,7 +2632,7 @@ restart_loop: memcpy(&tmp_from, &saved_from, sizeof(struct iov_iter)); iov_iter_advance(&tmp_from, - wdata->offset - *poffset); + wdata->offset - iocb->ki_pos); rc = cifs_write_from_iter(wdata->offset, wdata->bytes, &tmp_from, @@ -2650,34 +2649,13 @@ restart_loop: kref_put(&wdata->refcount, cifs_uncached_writedata_release); } - if (total_written > 0) - *poffset += total_written; + if (unlikely(!total_written)) + return rc; + iocb->ki_pos += total_written; + set_bit(CIFS_INO_INVALID_MAPPING, &CIFS_I(file_inode(file))->flags); cifs_stats_bytes_written(tcon, total_written); - return total_written ? total_written : (ssize_t)rc; -} - -ssize_t cifs_user_writev(struct kiocb *iocb, struct iov_iter *from) -{ - ssize_t written; - struct inode *inode; - loff_t pos = iocb->ki_pos; - - inode = file_inode(iocb->ki_filp); - - /* - * BB - optimize the way when signing is disabled. We can drop this - * extra memory-to-memory copying and use iovec buffers for constructing - * write request. - */ - - written = cifs_iovec_write(iocb->ki_filp, from, &pos); - if (written > 0) { - set_bit(CIFS_INO_INVALID_MAPPING, &CIFS_I(inode)->flags); - iocb->ki_pos = pos; - } - - return written; + return total_written; } static ssize_t @@ -2688,8 +2666,7 @@ cifs_writev(struct kiocb *iocb, struct iov_iter *from) struct inode *inode = file->f_mapping->host; struct cifsInodeInfo *cinode = CIFS_I(inode); struct TCP_Server_Info *server = tlink_tcon(cfile->tlink)->ses->server; - ssize_t rc = -EACCES; - loff_t lock_pos = iocb->ki_pos; + ssize_t rc; /* * We need to hold the sem to be sure nobody modifies lock list @@ -2697,23 +2674,24 @@ cifs_writev(struct kiocb *iocb, struct iov_iter *from) */ down_read(&cinode->lock_sem); mutex_lock(&inode->i_mutex); - if (file->f_flags & O_APPEND) - lock_pos = i_size_read(inode); - if (!cifs_find_lock_conflict(cfile, lock_pos, iov_iter_count(from), + + rc = generic_write_checks(iocb, from); + if (rc <= 0) + goto out; + + if (!cifs_find_lock_conflict(cfile, iocb->ki_pos, iov_iter_count(from), server->vals->exclusive_lock_type, NULL, - CIFS_WRITE_OP)) { + CIFS_WRITE_OP)) rc = __generic_file_write_iter(iocb, from); - mutex_unlock(&inode->i_mutex); - - if (rc > 0) { - ssize_t err; + else + rc = -EACCES; +out: + mutex_unlock(&inode->i_mutex); - err = generic_write_sync(file, iocb->ki_pos - rc, rc); - if (err < 0) - rc = err; - } - } else { - mutex_unlock(&inode->i_mutex); + if (rc > 0) { + ssize_t err = generic_write_sync(file, iocb->ki_pos - rc, rc); + if (err < 0) + rc = err; } up_read(&cinode->lock_sem); return rc; @@ -3877,8 +3855,7 @@ void cifs_oplock_break(struct work_struct *work) * Direct IO is not yet supported in the cached mode. */ static ssize_t -cifs_direct_io(int rw, struct kiocb *iocb, struct iov_iter *iter, - loff_t pos) +cifs_direct_io(struct kiocb *iocb, struct iov_iter *iter, loff_t pos) { /* * FIXME diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index cf0db005d2f5..acb3d63bc9dc 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -1598,7 +1598,7 @@ static loff_t configfs_dir_lseek(struct file *file, loff_t offset, int whence) if (offset >= 0) break; default: - mutex_unlock(&file_inode(file)->i_mutex); + mutex_unlock(&dentry->d_inode->i_mutex); return -EINVAL; } if (offset != file->f_pos) { @@ -98,9 +98,9 @@ static bool buffer_size_valid(struct buffer_head *bh) return bh->b_state != 0; } -static ssize_t dax_io(int rw, struct inode *inode, struct iov_iter *iter, - loff_t start, loff_t end, get_block_t get_block, - struct buffer_head *bh) +static ssize_t dax_io(struct inode *inode, struct iov_iter *iter, + loff_t start, loff_t end, get_block_t get_block, + struct buffer_head *bh) { ssize_t retval = 0; loff_t pos = start; @@ -109,7 +109,7 @@ static ssize_t dax_io(int rw, struct inode *inode, struct iov_iter *iter, void *addr; bool hole = false; - if (rw != WRITE) + if (iov_iter_rw(iter) != WRITE) end = min(end, i_size_read(inode)); while (pos < end) { @@ -124,7 +124,7 @@ static ssize_t dax_io(int rw, struct inode *inode, struct iov_iter *iter, bh->b_size = PAGE_ALIGN(end - pos); bh->b_state = 0; retval = get_block(inode, block, bh, - rw == WRITE); + iov_iter_rw(iter) == WRITE); if (retval) break; if (!buffer_size_valid(bh)) @@ -137,7 +137,7 @@ static ssize_t dax_io(int rw, struct inode *inode, struct iov_iter *iter, bh->b_size -= done; } - hole = (rw != WRITE) && !buffer_written(bh); + hole = iov_iter_rw(iter) != WRITE && !buffer_written(bh); if (hole) { addr = NULL; size = bh->b_size - first; @@ -154,7 +154,7 @@ static ssize_t dax_io(int rw, struct inode *inode, struct iov_iter *iter, max = min(pos + size, end); } - if (rw == WRITE) + if (iov_iter_rw(iter) == WRITE) len = copy_from_iter(addr, max - pos, iter); else if (!hole) len = copy_to_iter(addr, max - pos, iter); @@ -173,7 +173,6 @@ static ssize_t dax_io(int rw, struct inode *inode, struct iov_iter *iter, /** * dax_do_io - Perform I/O to a DAX file - * @rw: READ to read or WRITE to write * @iocb: The control block for this I/O * @inode: The file which the I/O is directed at * @iter: The addresses to do I/O from or to @@ -189,9 +188,9 @@ static ssize_t dax_io(int rw, struct inode *inode, struct iov_iter *iter, * As with do_blockdev_direct_IO(), we increment i_dio_count while the I/O * is in progress. */ -ssize_t dax_do_io(int rw, struct kiocb *iocb, struct inode *inode, - struct iov_iter *iter, loff_t pos, - get_block_t get_block, dio_iodone_t end_io, int flags) +ssize_t dax_do_io(struct kiocb *iocb, struct inode *inode, + struct iov_iter *iter, loff_t pos, get_block_t get_block, + dio_iodone_t end_io, int flags) { struct buffer_head bh; ssize_t retval = -EINVAL; @@ -199,7 +198,7 @@ ssize_t dax_do_io(int rw, struct kiocb *iocb, struct inode *inode, memset(&bh, 0, sizeof(bh)); - if ((flags & DIO_LOCKING) && (rw == READ)) { + if ((flags & DIO_LOCKING) && iov_iter_rw(iter) == READ) { struct address_space *mapping = inode->i_mapping; mutex_lock(&inode->i_mutex); retval = filemap_write_and_wait_range(mapping, pos, end - 1); @@ -212,9 +211,9 @@ ssize_t dax_do_io(int rw, struct kiocb *iocb, struct inode *inode, /* Protects against truncate */ atomic_inc(&inode->i_dio_count); - retval = dax_io(rw, inode, iter, pos, end, get_block, &bh); + retval = dax_io(inode, iter, pos, end, get_block, &bh); - if ((flags & DIO_LOCKING) && (rw == READ)) + if ((flags & DIO_LOCKING) && iov_iter_rw(iter) == READ) mutex_unlock(&inode->i_mutex); if ((retval > 0) && end_io) diff --git a/fs/dcache.c b/fs/dcache.c index d99736a63e3c..656ce522a218 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -269,6 +269,41 @@ static inline int dname_external(const struct dentry *dentry) return dentry->d_name.name != dentry->d_iname; } +/* + * Make sure other CPUs see the inode attached before the type is set. + */ +static inline void __d_set_inode_and_type(struct dentry *dentry, + struct inode *inode, + unsigned type_flags) +{ + unsigned flags; + + dentry->d_inode = inode; + smp_wmb(); + flags = READ_ONCE(dentry->d_flags); + flags &= ~(DCACHE_ENTRY_TYPE | DCACHE_FALLTHRU); + flags |= type_flags; + WRITE_ONCE(dentry->d_flags, flags); +} + +/* + * Ideally, we want to make sure that other CPUs see the flags cleared before + * the inode is detached, but this is really a violation of RCU principles + * since the ordering suggests we should always set inode before flags. + * + * We should instead replace or discard the entire dentry - but that sucks + * performancewise on mass deletion/rename. + */ +static inline void __d_clear_type_and_inode(struct dentry *dentry) +{ + unsigned flags = READ_ONCE(dentry->d_flags); + + flags &= ~(DCACHE_ENTRY_TYPE | DCACHE_FALLTHRU); + WRITE_ONCE(dentry->d_flags, flags); + smp_wmb(); + dentry->d_inode = NULL; +} + static void dentry_free(struct dentry *dentry) { WARN_ON(!hlist_unhashed(&dentry->d_u.d_alias)); @@ -311,7 +346,7 @@ static void dentry_iput(struct dentry * dentry) { struct inode *inode = dentry->d_inode; if (inode) { - dentry->d_inode = NULL; + __d_clear_type_and_inode(dentry); hlist_del_init(&dentry->d_u.d_alias); spin_unlock(&dentry->d_lock); spin_unlock(&inode->i_lock); @@ -335,8 +370,7 @@ static void dentry_unlink_inode(struct dentry * |