summaryrefslogtreecommitdiff
path: root/fs/f2fs/file.c
diff options
context:
space:
mode:
Diffstat (limited to 'fs/f2fs/file.c')
-rw-r--r--fs/f2fs/file.c662
1 files changed, 490 insertions, 172 deletions
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index 9c8ef33bd8d3..5b89af0f27f0 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -24,6 +24,7 @@
#include <linux/sched/signal.h>
#include <linux/fileattr.h>
#include <linux/fadvise.h>
+#include <linux/iomap.h>
#include "f2fs.h"
#include "node.h"
@@ -236,13 +237,13 @@ static void try_to_fix_pino(struct inode *inode)
struct f2fs_inode_info *fi = F2FS_I(inode);
nid_t pino;
- down_write(&fi->i_sem);
+ f2fs_down_write(&fi->i_sem);
if (file_wrong_pino(inode) && inode->i_nlink == 1 &&
get_parent_ino(inode, &pino)) {
f2fs_i_pino_write(inode, pino);
file_got_pino(inode);
}
- up_write(&fi->i_sem);
+ f2fs_up_write(&fi->i_sem);
}
static int f2fs_do_sync_file(struct file *file, loff_t start, loff_t end,
@@ -317,9 +318,9 @@ go_write:
* Both of fdatasync() and fsync() are able to be recovered from
* sudden-power-off.
*/
- down_read(&F2FS_I(inode)->i_sem);
+ f2fs_down_read(&F2FS_I(inode)->i_sem);
cp_reason = need_do_checkpoint(inode);
- up_read(&F2FS_I(inode)->i_sem);
+ f2fs_up_read(&F2FS_I(inode)->i_sem);
if (cp_reason) {
/* all the dirty node pages should be flushed for POR */
@@ -786,7 +787,7 @@ int f2fs_truncate(struct inode *inode)
return -EIO;
}
- err = dquot_initialize(inode);
+ err = f2fs_dquot_initialize(inode);
if (err)
return err;
@@ -811,7 +812,7 @@ int f2fs_getattr(struct user_namespace *mnt_userns, const struct path *path,
{
struct inode *inode = d_inode(path->dentry);
struct f2fs_inode_info *fi = F2FS_I(inode);
- struct f2fs_inode *ri;
+ struct f2fs_inode *ri = NULL;
unsigned int flags;
if (f2fs_has_extra_attr(inode) &&
@@ -843,7 +844,7 @@ int f2fs_getattr(struct user_namespace *mnt_userns, const struct path *path,
STATX_ATTR_NODUMP |
STATX_ATTR_VERITY);
- generic_fillattr(&init_user_ns, inode, stat);
+ generic_fillattr(mnt_userns, inode, stat);
/* we need to show initial sectors used for inline_data/dentries */
if ((S_ISREG(inode->i_mode) && f2fs_has_inline_data(inode)) ||
@@ -903,7 +904,7 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
!f2fs_is_compress_backend_ready(inode))
return -EOPNOTSUPP;
- err = setattr_prepare(&init_user_ns, dentry, attr);
+ err = setattr_prepare(mnt_userns, dentry, attr);
if (err)
return err;
@@ -916,7 +917,7 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
return err;
if (is_quota_modification(inode, attr)) {
- err = dquot_initialize(inode);
+ err = f2fs_dquot_initialize(inode);
if (err)
return err;
}
@@ -957,7 +958,7 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
return err;
}
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
filemap_invalidate_lock(inode->i_mapping);
truncate_setsize(inode, attr->ia_size);
@@ -969,7 +970,7 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
* larger than i_size.
*/
filemap_invalidate_unlock(inode->i_mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
if (err)
return err;
@@ -979,10 +980,10 @@ int f2fs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
spin_unlock(&F2FS_I(inode)->i_size_lock);
}
- __setattr_copy(&init_user_ns, inode, attr);
+ __setattr_copy(mnt_userns, inode, attr);
if (attr->ia_valid & ATTR_MODE) {
- err = posix_acl_chmod(&init_user_ns, inode, f2fs_get_inode_mode(inode));
+ err = posix_acl_chmod(mnt_userns, inode, f2fs_get_inode_mode(inode));
if (is_inode_flag_set(inode, FI_ACL_MODE)) {
if (!err)
@@ -1111,7 +1112,7 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len)
blk_start = (loff_t)pg_start << PAGE_SHIFT;
blk_end = (loff_t)pg_end << PAGE_SHIFT;
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
filemap_invalidate_lock(inode->i_mapping);
truncate_pagecache_range(inode, blk_start, blk_end - 1);
@@ -1121,7 +1122,7 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len)
f2fs_unlock_op(sbi);
filemap_invalidate_unlock(inode->i_mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
}
}
@@ -1232,7 +1233,7 @@ static int __clone_blkaddrs(struct inode *src_inode, struct inode *dst_inode,
if (ret)
return ret;
- ret = f2fs_get_node_info(sbi, dn.nid, &ni);
+ ret = f2fs_get_node_info(sbi, dn.nid, &ni, false);
if (ret) {
f2fs_put_dnode(&dn);
return ret;
@@ -1354,7 +1355,7 @@ static int f2fs_do_collapse(struct inode *inode, loff_t offset, loff_t len)
f2fs_balance_fs(sbi, true);
/* avoid gc operation during block exchange */
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
filemap_invalidate_lock(inode->i_mapping);
f2fs_lock_op(sbi);
@@ -1364,7 +1365,7 @@ static int f2fs_do_collapse(struct inode *inode, loff_t offset, loff_t len)
f2fs_unlock_op(sbi);
filemap_invalidate_unlock(inode->i_mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
return ret;
}
@@ -1499,7 +1500,7 @@ static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len,
unsigned int end_offset;
pgoff_t end;
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
filemap_invalidate_lock(mapping);
truncate_pagecache_range(inode,
@@ -1513,7 +1514,7 @@ static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len,
if (ret) {
f2fs_unlock_op(sbi);
filemap_invalidate_unlock(mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
goto out;
}
@@ -1525,7 +1526,7 @@ static int f2fs_zero_range(struct inode *inode, loff_t offset, loff_t len,
f2fs_unlock_op(sbi);
filemap_invalidate_unlock(mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
f2fs_balance_fs(sbi, dn.node_changed);
@@ -1599,7 +1600,7 @@ static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len)
idx = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE);
/* avoid gc operation during block exchange */
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
filemap_invalidate_lock(mapping);
truncate_pagecache(inode, offset);
@@ -1617,7 +1618,7 @@ static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len)
f2fs_unlock_op(sbi);
}
filemap_invalidate_unlock(mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
/* write out all moved pages, if possible */
filemap_invalidate_lock(mapping);
@@ -1673,13 +1674,13 @@ static int expand_inode_data(struct inode *inode, loff_t offset,
next_alloc:
if (has_not_enough_free_secs(sbi, 0,
GET_SEC_FROM_SEG(sbi, overprovision_segments(sbi)))) {
- down_write(&sbi->gc_lock);
+ f2fs_down_write(&sbi->gc_lock);
err = f2fs_gc(sbi, true, false, false, NULL_SEGNO);
if (err && err != -ENODATA && err != -EAGAIN)
goto out_err;
}
- down_write(&sbi->pin_sem);
+ f2fs_down_write(&sbi->pin_sem);
f2fs_lock_op(sbi);
f2fs_allocate_new_section(sbi, CURSEG_COLD_DATA_PINNED, false);
@@ -1687,8 +1688,9 @@ next_alloc:
map.m_seg_type = CURSEG_COLD_DATA_PINNED;
err = f2fs_map_blocks(inode, &map, 1, F2FS_GET_BLOCK_PRE_DIO);
+ file_dont_truncate(inode);
- up_write(&sbi->pin_sem);
+ f2fs_up_write(&sbi->pin_sem);
expanded += map.m_len;
sec_len -= map.m_len;
@@ -1748,7 +1750,11 @@ static long f2fs_fallocate(struct file *file, int mode,
(mode & (FALLOC_FL_COLLAPSE_RANGE | FALLOC_FL_INSERT_RANGE)))
return -EOPNOTSUPP;
- if (f2fs_compressed_file(inode) &&
+ /*
+ * Pinned file should not support partial trucation since the block
+ * can be used by applications.
+ */
+ if ((f2fs_compressed_file(inode) || f2fs_is_pinned_file(inode)) &&
(mode & (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_COLLAPSE_RANGE |
FALLOC_FL_ZERO_RANGE | FALLOC_FL_INSERT_RANGE)))
return -EOPNOTSUPP;
@@ -1983,11 +1989,12 @@ static int f2fs_ioc_getversion(struct file *filp, unsigned long arg)
static int f2fs_ioc_start_atomic_write(struct file *filp)
{
struct inode *inode = file_inode(filp);
+ struct user_namespace *mnt_userns = file_mnt_user_ns(filp);
struct f2fs_inode_info *fi = F2FS_I(inode);
struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
int ret;
- if (!inode_owner_or_capable(&init_user_ns, inode))
+ if (!inode_owner_or_capable(mnt_userns, inode))
return -EACCES;
if (!S_ISREG(inode->i_mode))
@@ -2002,7 +2009,10 @@ static int f2fs_ioc_start_atomic_write(struct file *filp)
inode_lock(inode);
- f2fs_disable_compressed_file(inode);
+ if (!f2fs_disable_compressed_file(inode)) {
+ ret = -EINVAL;
+ goto out;
+ }
if (f2fs_is_atomic_file(inode)) {
if (is_inode_flag_set(inode, FI_ATOMIC_REVOKE_REQUEST))
@@ -2014,7 +2024,7 @@ static int f2fs_ioc_start_atomic_write(struct file *filp)
if (ret)
goto out;
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
/*
* Should wait end_io to count F2FS_WB_CP_DATA correctly by
@@ -2025,7 +2035,7 @@ static int f2fs_ioc_start_atomic_write(struct file *filp)
inode->i_ino, get_dirty_pages(inode));
ret = filemap_write_and_wait_range(inode->i_mapping, 0, LLONG_MAX);
if (ret) {
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
goto out;
}
@@ -2038,7 +2048,7 @@ static int f2fs_ioc_start_atomic_write(struct file *filp)
/* add inode in inmem_list first and set atomic_file */
set_inode_flag(inode, FI_ATOMIC_FILE);
clear_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
f2fs_update_time(F2FS_I_SB(inode), REQ_TIME);
F2FS_I(inode)->inmem_task = current;
@@ -2052,9 +2062,10 @@ out:
static int f2fs_ioc_commit_atomic_write(struct file *filp)
{
struct inode *inode = file_inode(filp);
+ struct user_namespace *mnt_userns = file_mnt_user_ns(filp);
int ret;
- if (!inode_owner_or_capable(&init_user_ns, inode))
+ if (!inode_owner_or_capable(mnt_userns, inode))
return -EACCES;
ret = mnt_want_write_file(filp);
@@ -2094,9 +2105,10 @@ err_out:
static int f2fs_ioc_start_volatile_write(struct file *filp)
{
struct inode *inode = file_inode(filp);
+ struct user_namespace *mnt_userns = file_mnt_user_ns(filp);
int ret;
- if (!inode_owner_or_capable(&init_user_ns, inode))
+ if (!inode_owner_or_capable(mnt_userns, inode))
return -EACCES;
if (!S_ISREG(inode->i_mode))
@@ -2129,9 +2141,10 @@ out:
static int f2fs_ioc_release_volatile_write(struct file *filp)
{
struct inode *inode = file_inode(filp);
+ struct user_namespace *mnt_userns = file_mnt_user_ns(filp);
int ret;
- if (!inode_owner_or_capable(&init_user_ns, inode))
+ if (!inode_owner_or_capable(mnt_userns, inode))
return -EACCES;
ret = mnt_want_write_file(filp);
@@ -2158,9 +2171,10 @@ out:
static int f2fs_ioc_abort_volatile_write(struct file *filp)
{
struct inode *inode = file_inode(filp);
+ struct user_namespace *mnt_userns = file_mnt_user_ns(filp);
int ret;
- if (!inode_owner_or_capable(&init_user_ns, inode))
+ if (!inode_owner_or_capable(mnt_userns, inode))
return -EACCES;
ret = mnt_want_write_file(filp);
@@ -2345,7 +2359,7 @@ static int f2fs_ioc_get_encryption_pwsalt(struct file *filp, unsigned long arg)
if (err)
return err;
- down_write(&sbi->sb_lock);
+ f2fs_down_write(&sbi->sb_lock);
if (uuid_is_nonzero(sbi->raw_super->encrypt_pw_salt))
goto got_it;
@@ -2364,7 +2378,7 @@ got_it:
16))
err = -EFAULT;
out_err:
- up_write(&sbi->sb_lock);
+ f2fs_up_write(&sbi->sb_lock);
mnt_drop_write_file(filp);
return err;
}
@@ -2441,12 +2455,12 @@ static int f2fs_ioc_gc(struct file *filp, unsigned long arg)
return ret;
if (!sync) {
- if (!down_write_trylock(&sbi->gc_lock)) {
+ if (!f2fs_down_write_trylock(&sbi->gc_lock)) {
ret = -EBUSY;
goto out;
}
} else {
- down_write(&sbi->gc_lock);
+ f2fs_down_write(&sbi->gc_lock);
}
ret = f2fs_gc(sbi, sync, true, false, NULL_SEGNO);
@@ -2477,12 +2491,12 @@ static int __f2fs_ioc_gc_range(struct file *filp, struct f2fs_gc_range *range)
do_more:
if (!range->sync) {
- if (!down_write_trylock(&sbi->gc_lock)) {
+ if (!f2fs_down_write_trylock(&sbi->gc_lock)) {
ret = -EBUSY;
goto out;
}
} else {
- down_write(&sbi->gc_lock);
+ f2fs_down_write(&sbi->gc_lock);
}
ret = f2fs_gc(sbi, range->sync, true, false,
@@ -2553,10 +2567,6 @@ static int f2fs_defragment_range(struct f2fs_sb_info *sbi,
bool fragmented = false;
int err;
- /* if in-place-update policy is enabled, don't waste time here */
- if (f2fs_should_update_inplace(inode, NULL))
- return -EINVAL;
-
pg_start = range->start >> PAGE_SHIFT;
pg_end = (range->start + range->len) >> PAGE_SHIFT;
@@ -2564,6 +2574,13 @@ static int f2fs_defragment_range(struct f2fs_sb_info *sbi,
inode_lock(inode);
+ /* if in-place-update policy is enabled, don't waste time here */
+ set_inode_flag(inode, FI_OPU_WRITE);
+ if (f2fs_should_update_inplace(inode, NULL)) {
+ err = -EINVAL;
+ goto out;
+ }
+
/* writeback all dirty pages in the range */
err = filemap_write_and_wait_range(inode->i_mapping, range->start,
range->start + range->len - 1);
@@ -2645,7 +2662,7 @@ do_map:
goto check;
}
- set_inode_flag(inode, FI_DO_DEFRAG);
+ set_inode_flag(inode, FI_SKIP_WRITES);
idx = map.m_lblk;
while (idx < map.m_lblk + map.m_len && cnt < blk_per_seg) {
@@ -2670,15 +2687,16 @@ check:
if (map.m_lblk < pg_end && cnt < blk_per_seg)
goto do_map;
- clear_inode_flag(inode, FI_DO_DEFRAG);
+ clear_inode_flag(inode, FI_SKIP_WRITES);
err = filemap_fdatawrite(inode->i_mapping);
if (err)
goto out;
}
clear_out:
- clear_inode_flag(inode, FI_DO_DEFRAG);
+ clear_inode_flag(inode, FI_SKIP_WRITES);
out:
+ clear_inode_flag(inode, FI_OPU_WRITE);
inode_unlock(inode);
if (!err)
range->len = (u64)total << PAGE_SHIFT;
@@ -2814,10 +2832,10 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in,
f2fs_balance_fs(sbi, true);
- down_write(&F2FS_I(src)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(src)->i_gc_rwsem[WRITE]);
if (src != dst) {
ret = -EBUSY;
- if (!down_write_trylock(&F2FS_I(dst)->i_gc_rwsem[WRITE]))
+ if (!f2fs_down_write_trylock(&F2FS_I(dst)->i_gc_rwsem[WRITE]))
goto out_src;
}
@@ -2835,9 +2853,9 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in,
f2fs_unlock_op(sbi);
if (src != dst)
- up_write(&F2FS_I(dst)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(dst)->i_gc_rwsem[WRITE]);
out_src:
- up_write(&F2FS_I(src)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(src)->i_gc_rwsem[WRITE]);
out_unlock:
if (src != dst)
inode_unlock(dst);
@@ -2932,7 +2950,7 @@ static int f2fs_ioc_flush_device(struct file *filp, unsigned long arg)
end_segno = min(start_segno + range.segments, dev_end_segno);
while (start_segno < end_segno) {
- if (!down_write_trylock(&sbi->gc_lock)) {
+ if (!f2fs_down_write_trylock(&sbi->gc_lock)) {
ret = -EBUSY;
goto out;
}
@@ -2984,7 +3002,7 @@ static int f2fs_ioc_setproject(struct inode *inode, __u32 projid)
{
struct f2fs_inode_info *fi = F2FS_I(inode);
struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
- struct page *ipage;
+ struct f2fs_inode *ri = NULL;
kprojid_t kprojid;
int err;
@@ -3008,19 +3026,10 @@ static int f2fs_ioc_setproject(struct inode *inode, __u32 projid)
if (IS_NOQUOTA(inode))
return err;
- ipage = f2fs_get_node_page(sbi, inode->i_ino);
- if (IS_ERR(ipage))
- return PTR_ERR(ipage);
+ if (!F2FS_FITS_IN_INODE(ri, fi->i_extra_isize, i_projid))
+ return -EOVERFLOW;
- if (!F2FS_FITS_IN_INODE(F2FS_INODE(ipage), fi->i_extra_isize,
- i_projid)) {
- err = -EOVERFLOW;
- f2fs_put_page(ipage, 1);
- return err;
- }
- f2fs_put_page(ipage, 1);
-
- err = dquot_initialize(inode);
+ err = f2fs_dquot_initialize(inode);
if (err)
return err;
@@ -3143,17 +3152,17 @@ static int f2fs_ioc_set_pin_file(struct file *filp, unsigned long arg)
inode_lock(inode);
- if (f2fs_should_update_outplace(inode, NULL)) {
- ret = -EINVAL;
- goto out;
- }
-
if (!pin) {
clear_inode_flag(inode, FI_PIN_FILE);
f2fs_i_gc_failures_write(inode, 0);
goto done;
}
+ if (f2fs_should_update_outplace(inode, NULL)) {
+ ret = -EINVAL;
+ goto out;
+ }
+
if (f2fs_pin_file_control(inode, false)) {
ret = -EAGAIN;
goto out;
@@ -3209,9 +3218,9 @@ int f2fs_precache_extents(struct inode *inode)
while (map.m_lblk < end) {
map.m_len = end - map.m_lblk;
- down_write(&fi->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&fi->i_gc_rwsem[WRITE]);
err = f2fs_map_blocks(inode, &map, 0, F2FS_GET_BLOCK_PRECACHE);
- up_write(&fi->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&fi->i_gc_rwsem[WRITE]);
if (err)
return err;
@@ -3288,11 +3297,11 @@ static int f2fs_ioc_getfslabel(struct file *filp, unsigned long arg)
if (!vbuf)
return -ENOMEM;
- down_read(&sbi->sb_lock);
+ f2fs_down_read(&sbi->sb_lock);
count = utf16s_to_utf8s(sbi->raw_super->volume_name,
ARRAY_SIZE(sbi->raw_super->volume_name),
UTF16_LITTLE_ENDIAN, vbuf, MAX_VOLUME_NAME);
- up_read(&sbi->sb_lock);
+ f2fs_up_read(&sbi->sb_lock);
if (copy_to_user((char __user *)arg, vbuf,
min(FSLABEL_MAX, count)))
@@ -3320,7 +3329,7 @@ static int f2fs_ioc_setfslabel(struct file *filp, unsigned long arg)
if (err)
goto out;
- down_write(&sbi->sb_lock);
+ f2fs_down_write(&sbi->sb_lock);
memset(sbi->raw_super->volume_name, 0,
sizeof(sbi->raw_super->volume_name));
@@ -3330,7 +3339,7 @@ static int f2fs_ioc_setfslabel(struct file *filp, unsigned long arg)
err = f2fs_commit_super(sbi, false);
- up_write(&sbi->sb_lock);
+ f2fs_up_write(&sbi->sb_lock);
mnt_drop_write_file(filp);
out:
@@ -3456,7 +3465,7 @@ static int f2fs_release_compress_blocks(struct file *filp, unsigned long arg)
if (!atomic_read(&F2FS_I(inode)->i_compr_blocks))
goto out;
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
filemap_invalidate_lock(inode->i_mapping);
last_idx = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE);
@@ -3493,7 +3502,7 @@ static int f2fs_release_compress_blocks(struct file *filp, unsigned long arg)
}
filemap_invalidate_unlock(inode->i_mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
out:
inode_unlock(inode);
@@ -3609,7 +3618,7 @@ static int f2fs_reserve_compress_blocks(struct file *filp, unsigned long arg)
goto unlock_inode;
}
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
filemap_invalidate_lock(inode->i_mapping);
last_idx = DIV_ROUND_UP(i_size_read(inode), PAGE_SIZE);
@@ -3646,7 +3655,7 @@ static int f2fs_reserve_compress_blocks(struct file *filp, unsigned long arg)
}
filemap_invalidate_unlock(inode->i_mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
if (ret >= 0) {
clear_inode_flag(inode, FI_COMPRESS_RELEASED);
@@ -3764,7 +3773,7 @@ static int f2fs_sec_trim_file(struct file *filp, unsigned long arg)
if (ret)
goto err;
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
filemap_invalidate_lock(mapping);
ret = filemap_write_and_wait_range(mapping, range.start,
@@ -3853,7 +3862,7 @@ static int f2fs_sec_trim_file(struct file *filp, unsigned long arg)
prev_block, len, range.flags);
out:
filemap_invalidate_unlock(mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
err:
inode_unlock(inode);
file_end_write(filp);
@@ -4218,133 +4227,436 @@ long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg)
return __f2fs_ioctl(filp, cmd, arg);
}
-static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter)
+/*
+ * Return %true if the given read or write request should use direct I/O, or
+ * %false if it should use buffered I/O.
+ */
+static bool f2fs_should_use_dio(struct inode *inode, struct kiocb *iocb,
+ struct iov_iter *iter)
+{
+ unsigned int align;
+
+ if (!(iocb->ki_flags & IOCB_DIRECT))
+ return false;
+
+ if (f2fs_force_buffered_io(inode, iocb, iter))
+ return false;
+
+ /*
+ * Direct I/O not aligned to the disk's logical_block_size will be
+ * attempted, but will fail with -EINVAL.
+ *
+ * f2fs additionally requires that direct I/O be aligned to the
+ * filesystem block size, which is often a stricter requirement.
+ * However, f2fs traditionally falls back to buffered I/O on requests
+ * that are logical_block_size-aligned but not fs-block aligned.
+ *
+ * The below logic implements this behavior.
+ */
+ align = iocb->ki_pos | iov_iter_alignment(iter);
+ if (!IS_ALIGNED(align, i_blocksize(inode)) &&
+ IS_ALIGNED(align, bdev_logical_block_size(inode->i_sb->s_bdev)))
+ return false;
+
+ return true;
+}
+
+static int f2fs_dio_read_end_io(struct kiocb *iocb, ssize_t size, int error,
+ unsigned int flags)
+{
+ struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(iocb->ki_filp));
+
+ dec_page_count(sbi, F2FS_DIO_READ);
+ if (error)
+ return error;
+ f2fs_update_iostat(sbi, APP_DIRECT_READ_IO, size);
+ return 0;
+}
+
+static const struct iomap_dio_ops f2fs_iomap_dio_read_ops = {
+ .end_io = f2fs_dio_read_end_io,
+};
+
+static ssize_t f2fs_dio_read_iter(struct kiocb *iocb, struct iov_iter *to)
{
struct file *file = iocb->ki_filp;
struct inode *inode = file_inode(file);
- int ret;
+ struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+ struct f2fs_inode_info *fi = F2FS_I(inode);
+ const loff_t pos = iocb->ki_pos;
+ const size_t count = iov_iter_count(to);
+ struct iomap_dio *dio;
+ ssize_t ret;
+
+ if (count == 0)
+ return 0; /* skip atime update */
+
+ trace_f2fs_direct_IO_enter(inode, iocb, count, READ);
+
+ if (iocb->ki_flags & IOCB_NOWAIT) {
+ if (!f2fs_down_read_trylock(&fi->i_gc_rwsem[READ])) {
+ ret = -EAGAIN;
+ goto out;
+ }
+ } else {
+ f2fs_down_read(&fi->i_gc_rwsem[READ]);
+ }
+
+ /*
+ * We have to use __iomap_dio_rw() and iomap_dio_complete() instead of
+ * the higher-level function iomap_dio_rw() in order to ensure that the
+ * F2FS_DIO_READ counter will be decremented correctly in all cases.
+ */
+ inc_page_count(sbi, F2FS_DIO_READ);
+ dio = __iomap_dio_rw(iocb, to, &f2fs_iomap_ops,
+ &f2fs_iomap_dio_read_ops, 0, 0);
+ if (IS_ERR_OR_NULL(dio)) {
+ ret = PTR_ERR_OR_ZERO(dio);
+ if (ret != -EIOCBQUEUED)
+ dec_page_count(sbi, F2FS_DIO_READ);
+ } else {
+ ret = iomap_dio_complete(dio);
+ }
+
+ f2fs_up_read(&fi->i_gc_rwsem[READ]);
+
+ file_accessed(file);
+out:
+ trace_f2fs_direct_IO_exit(inode, pos, count, READ, ret);
+ return ret;
+}
+
+static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
+{
+ struct inode *inode = file_inode(iocb->ki_filp);
+ ssize_t ret;
if (!f2fs_is_compress_backend_ready(inode))
return -EOPNOTSUPP;
- ret = generic_file_read_iter(iocb, iter);
+ if (f2fs_should_use_dio(inode, iocb, to))
+ return f2fs_dio_read_iter(iocb, to);
+ ret = filemap_read(iocb, to, 0);
if (ret > 0)
- f2fs_update_iostat(F2FS_I_SB(inode), APP_READ_IO, ret);
-
+ f2fs_update_iostat(F2FS_I_SB(inode), APP_BUFFERED_READ_IO, ret);
return ret;
}
-static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
+static ssize_t f2fs_write_checks(struct kiocb *iocb, struct iov_iter *from)
{
struct file *file = iocb->ki_filp;
struct inode *inode = file_inode(file);
- ssize_t ret;
+ ssize_t count;
+ int err;
- if (unlikely(f2fs_cp_error(F2FS_I_SB(inode)))) {
- ret = -EIO;
- goto out;
+ if (IS_IMMUTABLE(inode))
+ return -EPERM;
+
+ if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED))
+ return -EPERM;
+
+ count = generic_write_checks(iocb, from);
+ if (count <= 0)
+ return count;
+
+ err = file_modified(file);
+ if (err)
+ return err;
+ return count;
+}
+
+/*
+ * Preallocate blocks for a write request, if it is possible and helpful to do
+ * so. Returns a positive number if blocks may have been preallocated, 0 if no
+ * blocks were preallocated, or a negative errno value if something went
+ * seriously wrong. Also sets FI_PREALLOCATED_ALL on the inode if *all* the
+ * requested blocks (not just some of them) have been allocated.
+ */
+static int f2fs_preallocate_blocks(struct kiocb *iocb, struct iov_iter *iter,
+ bool dio)
+{
+ struct inode *inode = file_inode(iocb->ki_filp);
+ struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+ const loff_t pos = iocb->ki_pos;
+ const size_t count = iov_iter_count(iter);
+ struct f2fs_map_blocks map = {};
+ int flag;
+ int ret;
+
+ /* If it will be an out-of-place direct write, don't bother. */
+ if (dio && f2fs_lfs_mode(sbi))
+ return 0;
+ /*
+ * Don't preallocate holes aligned to DIO_SKIP_HOLES which turns into
+ * buffered IO, if DIO meets any holes.
+ */
+ if (dio && i_size_read(inode) &&
+ (F2FS_BYTES_TO_BLK(pos) < F2FS_BLK_ALIGN(i_size_read(inode))))
+ return 0;
+
+ /* No-wait I/O can't allocate blocks. */
+ if (iocb->ki_flags & IOCB_NOWAIT)
+ return 0;
+
+ /* If it will be a short write, don't bother. */
+ if (fault_in_iov_iter_readable(iter, count))
+ return 0;
+
+ if (f2fs_has_inline_data(inode)) {
+ /* If the data will fit inline, don't bother. */
+ if (pos + count <= MAX_INLINE_DATA(inode))
+ return 0;
+ ret = f2fs_convert_inline_inode(inode);
+ if (ret)
+ return ret;
}
- if (!f2fs_is_compress_backend_ready(inode)) {
- ret = -EOPNOTSUPP;
- goto out;
+ /* Do not preallocate blocks that will be written partially in 4KB. */
+ map.m_lblk = F2FS_BLK_ALIGN(pos);
+ map.m_len = F2FS_BYTES_TO_BLK(pos + count);
+ if (map.m_len > map.m_lblk)
+ map.m_len -= map.m_lblk;
+ else
+ map.m_len = 0;
+ map.m_may_create = true;
+ if (dio) {
+ map.m_seg_type = f2fs_rw_hint_to_seg_type(inode->i_write_hint);
+ flag = F2FS_GET_BLOCK_PRE_DIO;
+ } else {
+ map.m_seg_type = NO_CHECK_TYPE;
+ flag = F2FS_GET_BLOCK_PRE_AIO;
}
+ ret = f2fs_map_blocks(inode, &map, 1, flag);
+ /* -ENOSPC|-EDQUOT are fine to report the number of allocated blocks. */
+ if (ret < 0 && !((ret == -ENOSPC || ret == -EDQUOT) && map.m_len > 0))
+ return ret;
+ if (ret == 0)
+ set_inode_flag(inode, FI_PREALLOCATED_ALL);
+ return map.m_len;
+}
+
+static ssize_t f2fs_buffered_write_iter(struct kiocb *iocb,
+ struct iov_iter *from)
+{
+ struct file *file = iocb->ki_filp;
+ struct inode *inode = file_inode(file);
+ ssize_t ret;
+
+ if (iocb->ki_flags & IOCB_NOWAIT)
+ return -EOPNOTSUPP;
+
+ current->backing_dev_info = inode_to_bdi(inode);
+ ret = generic_perform_write(iocb, from);
+ current->backing_dev_info = NULL;
+
+ if (ret > 0) {
+ iocb->ki_pos += ret;
+ f2fs_update_iostat(F2FS_I_SB(inode), APP_BUFFERED_IO, ret);
+ }
+ return ret;
+}
+
+static int f2fs_dio_write_end_io(struct kiocb *iocb, ssize_t size, int error,
+ unsigned int flags)
+{
+ struct f2fs_sb_info *sbi = F2FS_I_SB(file_inode(iocb->ki_filp));
+
+ dec_page_count(sbi, F2FS_DIO_WRITE);
+ if (error)
+ return error;
+ f2fs_update_iostat(sbi, APP_DIRECT_IO, size);
+ return 0;
+}
+
+static const struct iomap_dio_ops f2fs_iomap_dio_write_ops = {
+ .end_io = f2fs_dio_write_end_io,
+};
+
+static ssize_t f2fs_dio_write_iter(struct kiocb *iocb, struct iov_iter *from,
+ bool *may_need_sync)
+{
+ struct file *file = iocb->ki_filp;
+ struct inode *inode = file_inode(file);
+ struct f2fs_inode_info *fi = F2FS_I(inode);
+ struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
+ const bool do_opu = f2fs_lfs_mode(sbi);
+ const loff_t pos = iocb->ki_pos;
+ const ssize_t count = iov_iter_count(from);
+ unsigned int dio_flags;
+ struct iomap_dio *dio;
+ ssize_t ret;
+
+ trace_f2fs_direct_IO_enter(inode, iocb, count, WRITE);
+
if (iocb->ki_flags & IOCB_NOWAIT) {
- if (!inode_trylock(inode)) {
+ /* f2fs_convert_inline_inode() and block allocation can block */
+ if (f2fs_has_inline_data(inode) ||
+ !f2fs_overwrite_io(inode, pos, count)) {
+ ret = -EAGAIN;
+ goto out;
+ }
+
+ if (!f2fs_down_read_trylock(&fi->i_gc_rwsem[WRITE])) {
+ ret = -EAGAIN;
+ goto out;
+ }
+ if (do_opu && !f2fs_down_read_trylock(&fi->i_gc_rwsem[READ])) {
+ f2fs_up_read(&fi->i_gc_rwsem[WRITE]);
ret = -EAGAIN;
goto out;
}
} else {
- inode_lock(inode);
- }
+ ret = f2fs_convert_inline_inode(inode);
+ if (ret)
+ goto out;
- if (unlikely(IS_IMMUTABLE(inode))) {
- ret = -EPERM;
- goto unlock;
+ f2fs_down_read(&fi->i_gc_rwsem[WRITE]);
+ if (do_opu)
+ f2fs_down_read(&fi->i_gc_rwsem[READ]);
}
- if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED)) {
- ret = -EPERM;
- goto unlock;
+ /*
+ * We have to use __iomap_dio_rw() and iomap_dio_complete() instead of
+ * the higher-level function iomap_dio_rw() in order to ensure that the
+ * F2FS_DIO_WRITE counter will be decremented correctly in all cases.
+ */
+ inc_page_count(sbi, F2FS_DIO_WRITE);
+ dio_flags = 0;
+ if (pos + count > inode->i_size)
+ dio_flags |= IOMAP_DIO_FORCE_WAIT;
+ dio = __iomap_dio_rw(iocb, from, &f2fs_iomap_ops,
+ &f2fs_iomap_dio_write_ops, dio_flags, 0);
+ if (IS_ERR_OR_NULL(dio)) {
+ ret = PTR_ERR_OR_ZERO(dio);
+ if (ret == -ENOTBLK)
+ ret = 0;
+ if (ret != -EIOCBQUEUED)
+ dec_page_count(sbi, F2FS_DIO_WRITE);
+ } else {
+ ret = iomap_dio_complete(dio);
}
- ret = generic_write_checks(iocb, from);
- if (ret > 0) {
- bool preallocated = false;
- size_t target_size = 0;
- int err;
-
- if (iov_iter_fault_in_readable(from, iov_iter_count(from)))
- set_inode_flag(inode, FI_NO_PREALLOC);
-
- if ((iocb->ki_flags & IOCB_NOWAIT)) {
- if (!f2fs_overwrite_io(inode, iocb->ki_pos,
- iov_iter_count(from)) ||
- f2fs_has_inline_data(inode) ||
- f2fs_force_buffered_io(inode, iocb, from)) {
- clear_inode_flag(inode, FI_NO_PREALLOC);
- inode_unlock(inode);
- ret = -EAGAIN;
+ if (do_opu)
+ f2fs_up_read(&fi->i_gc_rwsem[READ]);
+ f2fs_up_read(&fi->i_gc_rwsem[WRITE]);
+
+ if (ret < 0)
+ goto out;
+ if (pos + ret > inode->i_size)
+ f2fs_i_size_write(inode, pos + ret);
+ if (!do_opu)
+ set_inode_flag(inode, FI_UPDATE_WRITE);
+
+ if (iov_iter_count(from)) {
+ ssize_t ret2;
+ loff_t bufio_start_pos = iocb->ki_pos;
+
+ /*
+ * The direct write was partial, so we need to fall back to a
+ * buffered write for the remainder.
+ */
+
+ ret2 = f2fs_buffered_write_iter(iocb, from);
+ if (iov_iter_count(from))
+ f2fs_write_failed(inode, iocb->ki_pos);
+ if (ret2 < 0)
+ goto out;
+
+ /*
+ * Ensure that the pagecache pages are written to disk and
+ * invalidated to preserve the expected O_DIRECT semantics.
+ */
+ if (ret2 > 0) {
+ loff_t bufio_end_pos = bufio_start_pos + ret2 - 1;
+
+ ret += ret2;
+
+ ret2 = filemap_write_and_wait_range(file->f_mapping,
+ bufio_start_pos,
+ bufio_end_pos);
+ if (ret2 < 0)
goto out;
- }
- goto write;
+ invalidate_mapping_pages(file->f_mapping,
+ bufio_start_pos >> PAGE_SHIFT,
+ bufio_end_pos >> PAGE_SHIFT);
}
+ } else {
+ /* iomap_dio_rw() already handled the generic_write_sync(). */
+ *may_need_sync = false;
+ }
+out:
+ trace_f2fs_direct_IO_exit(inode, pos, count, WRITE, ret);
+ return ret;
+}
- if (is_inode_flag_set(inode, FI_NO_PREALLOC))
- goto write;
+static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from)
+{
+ struct inode *inode = file_inode(iocb->ki_filp);
+ const loff_t orig_pos = iocb->ki_pos;
+ const size_t orig_count = iov_iter_count(from);
+ loff_t target_size;
+ bool dio;
+ bool may_need_sync = true;
+ int preallocated;
+ ssize_t ret;
- if (iocb->ki_flags & IOCB_DIRECT) {
- /*
- * Convert inline data for Direct I/O before entering
- * f2fs_direct_IO().
- */
- err = f2fs_convert_inline_inode(inode);
- if (err)
- goto out_err;
- /*
- * If force_buffere_io() is true, we have to allocate
- * blocks all the time, since f2fs_direct_IO will fall
- * back to buffered IO.
- */
- if (!f2fs_force_buffered_io(inode, iocb, from) &&
- f2fs_lfs_mode(F2FS_I_SB(inode)))
- goto write;
- }
- preallocated = true;
- target_size = iocb->ki_pos + iov_iter_count(from);
+ if (unlikely(f2fs_cp_error(F2FS_I_SB(inode)))) {
+ ret = -EIO;
+ goto out;
+ }
- err = f2fs_preallocate_blocks(iocb, from);
- if (err) {
-out_err:
- clear_inode_flag(inode, FI_NO_PREALLOC);
- inode_unlock(inode);
- ret = err;
+ if (!f2fs_is_compress_backend_ready(inode)) {
+ ret = -EOPNOTSUPP;
+ goto out;
+ }
+
+ if (iocb->ki_flags & IOCB_NOWAIT) {
+ if (!inode_trylock(inode)) {
+ ret = -EAGAIN;
goto out;
}
-write:
- ret = __generic_file_write_iter(iocb, from);
- clear_inode_flag(inode, FI_NO_PREALLOC);
+ } else {
+ inode_lock(inode);
+ }
- /* if we couldn't write data, we should deallocate blocks. */
- if (preallocated && i_size_read(inode) < target_size) {
- down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
- filemap_invalidate_lock(inode->i_mapping);
- f2fs_truncate(inode);
- filemap_invalidate_unlock(inode->i_mapping);
- up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
- }
+ ret = f2fs_write_checks(iocb, from);
+ if (ret <= 0)
+ goto out_unlock;
- if (ret > 0)
- f2fs_update_iostat(F2FS_I_SB(inode), APP_WRITE_IO, ret);
+ /* Determine whether we will do a direct write or a buffered write. */
+ dio = f2fs_should_use_dio(inode, iocb, from);
+
+ /* Possibly preallocate the blocks for the write. */
+ target_size = iocb->ki_pos + iov_iter_count(from);
+ preallocated = f2fs_preallocate_blocks(iocb, from, dio);
+ if (preallocated < 0)
+ ret = preallocated;
+ else
+ /* Do the actual write. */
+ ret = dio ?
+ f2fs_dio_write_iter(iocb, from, &may_need_sync):
+ f2fs_buffered_write_iter(iocb, from);
+
+ /* Don't leave any preallocated blocks around past i_size. */
+ if (preallocated && i_size_read(inode) < target_size) {
+ f2fs_down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ filemap_invalidate_lock(inode->i_mapping);
+ if (!f2fs_truncate(inode))
+ file_dont_truncate(inode);
+ filemap_invalidate_unlock(inode->i_mapping);
+ f2fs_up_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
+ } else {
+ file_dont_truncate(inode);
}
-unlock:
+
+ clear_inode_flag(inode, FI_PREALLOCATED_ALL);
+out_unlock:
inode_unlock(inode);
out:
- trace_f2fs_file_write_iter(inode, iocb->ki_pos,
- iov_iter_count(from), ret);
- if (ret > 0)
+ trace_f2fs_file_write_iter(inode, orig_pos, orig_count, ret);
+ if (ret > 0 && may_need_sync)
ret = generic_write_sync(iocb, ret);
return ret;
}
@@ -4352,12 +4664,12 @@ out:
static int f2fs_file_fadvise(struct file *filp, loff_t offset, loff_t len,
int advice)
{
- struct inode *inode;
struct address_space *mapping;
struct backing_dev_info *bdi;
+ struct inode *inode = file_inode(filp);
+ int err;
if (advice == POSIX_FADV_SEQUENTIAL) {
- inode = file_inode(filp);
if (S_ISFIFO(inode->i_mode))
return -ESPIPE;
@@ -4374,7 +4686,13 @@ static int f2fs_file_fadvise(struct file *filp, loff_t offset, loff_t len,
return 0;
}
- return generic_fadvise(filp, offset, len, advice);
+ err = generic_fadvise(filp, offset, len, advice);
+ if (!err && advice == POSIX_FADV_DONTNEED &&
+ test_opt(F2FS_I_SB(inode), COMPRESS_CACHE) &&
+ f2fs_compressed_file(inode))
+ f2fs_invalidate_compress_pages(F2FS_I_SB(inode), inode->i_ino);
+
+ return err;
}
#ifdef CONFIG_COMPAT
generated by cgit v1.2.3 (git 2.39.0) at 2024-09-06 16:12:47 +0300