[PATCH v2 11/20] f2fs: support subpage blocks in move range

From: Daeho Jeong

Date: Sun Nov 15 2026 - 23:10:08 EST


From: Daeho Jeong <daehojeong@xxxxxxxxxx>

Folios at both ends of the moved ranges can straddle the range
boundaries. Write back and truncate the page cache from PAGE_SIZE
boundaries, and hold invalidate_lock of both inodes during the exchange
so that faults and readahead cannot bring back stale data. After the
exchange, also after one that failed part way, drop the source blocks
that it read into the page cache, and call pagecache_isize_extended()
for the gap between the old EOF of dst and the moved data. The page
cache was not truncated before, so this also applies when the block
size equals PAGE_SIZE. Remove the -EOPNOTSUPP check.

Zero the cached data past the old EOF of dst with
f2fs_zero_post_eof_page() before the move extends dst over it, as the
other paths that extend a file do. Otherwise data stored past EOF
through a mapping becomes file data. This also applies when the block
size equals PAGE_SIZE.

Signed-off-by: Daeho Jeong <daehojeong@xxxxxxxxxx>
---
fs/f2fs/file.c | 73 +++++++++++++++++++++++++++++++++++++++-----------
1 file changed, 58 insertions(+), 15 deletions(-)

diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index d4472143dea0..4499a461be86 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -3435,7 +3435,7 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in,
size_t olen = len;
loff_t dst_max_i_size = 0;
loff_t dst_osize, dst_end;
- int ret;
+ int ret, err;

if (file_in->f_path.mnt != file_out->f_path.mnt ||
src->i_sb != dst->i_sb)
@@ -3516,17 +3516,6 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in,
if (ret)
goto out_unlock;

- /* write out all dirty pages from offset */
- ret = filemap_write_and_wait_range(src->i_mapping,
- pos_in, pos_in + len);
- if (ret)
- goto out_unlock;
-
- ret = filemap_write_and_wait_range(dst->i_mapping,
- pos_out, pos_out + len);
- if (ret)
- goto out_unlock;
-
f2fs_balance_fs(sbi, true);

f2fs_down_write(&F2FS_I(src)->i_gc_rwsem[WRITE]);
@@ -3536,6 +3525,39 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in,
goto out_src;
}

+ /*
+ * Hold invalidate_lock across writeback, truncation and the block
+ * exchange so neither faults nor readahead can repopulate the page
+ * cache with pre-exchange data.
+ */
+ filemap_invalidate_lock_two(src->i_mapping, dst->i_mapping);
+
+ /*
+ * Zero the cached data past the old EOF of dst, such as mmap stores
+ * past EOF, before the move extends dst over it.
+ */
+ ret = f2fs_zero_post_eof_page(dst, dst_end, false, false);
+ if (ret)
+ goto out_inval;
+
+ /* write out all dirty pages from offset */
+ ret = filemap_write_and_wait_range(src->i_mapping,
+ round_down(pos_in, PAGE_SIZE),
+ round_up(pos_in + len, PAGE_SIZE) - 1);
+ if (ret)
+ goto out_inval;
+
+ ret = filemap_write_and_wait_range(dst->i_mapping,
+ round_down(pos_out, PAGE_SIZE),
+ round_up(pos_out + len, PAGE_SIZE) - 1);
+ if (ret)
+ goto out_inval;
+
+ truncate_pagecache_range(src, round_down(pos_in, PAGE_SIZE),
+ round_up(pos_in + len, PAGE_SIZE) - 1);
+ truncate_pagecache_range(dst, round_down(pos_out, PAGE_SIZE),
+ round_up(pos_out + len, PAGE_SIZE) - 1);
+
f2fs_lock_op(sbi, &lc);
ret = __exchange_data_block(src, dst, F2FS_BYTES_TO_BLK(sbi, pos_in),
F2FS_BYTES_TO_BLK(sbi, pos_out),
@@ -3549,6 +3571,30 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in,
}
f2fs_unlock_op(sbi, &lc);

+ /*
+ * __clone_blkaddrs() reads the checkpointed source blocks into the
+ * page cache before punching them, so drop them again. Do it after a
+ * failed exchange too: the blocks moved before the failure were
+ * punched as well.
+ */
+ err = filemap_write_and_wait_range(src->i_mapping,
+ round_down(pos_in, PAGE_SIZE),
+ round_up(pos_in + len, PAGE_SIZE) - 1);
+ if (!err)
+ truncate_pagecache_range(src, round_down(pos_in, PAGE_SIZE),
+ round_up(pos_in + len, PAGE_SIZE) - 1);
+ if (!ret)
+ ret = err;
+
+ /*
+ * Write-protect the folio straddling the old EOF of dst, so that mmap
+ * stores to the blocks between it and the moved data fault.
+ */
+ if (i_size_read(dst) > dst_osize)
+ pagecache_isize_extended(dst, dst_osize, pos_out);
+out_inval:
+ filemap_invalidate_unlock_two(src->i_mapping, dst->i_mapping);
+
if (src != dst)
f2fs_up_write(&F2FS_I(dst)->i_gc_rwsem[WRITE]);
out_src:
@@ -3603,9 +3649,6 @@ static int f2fs_ioc_move_range(struct file *filp, unsigned long arg)
{
struct f2fs_move_range range;

- if (f2fs_has_subpage_blocks(F2FS_I_SB(file_inode(filp))))
- return -EOPNOTSUPP;
-
if (copy_from_user(&range, (struct f2fs_move_range __user *)arg,
sizeof(range)))
return -EFAULT;
--
2.56.0.611.g1030f0b940-goog