Blocks are tracked by u32, so the max permitted filesize is
U32_MAX * BLOCK_SIZE. Additionally, in order to support crypto block
sizes of 4K with a 16K block size, we must further restrict max filesize
to U32_MAX * 4096. This does not affect 4K blocksize f2fs as the natural
limits for those are well below that limit.

Signed-off-by: Daniel Rosenberg <drosen@google.com>
---
 fs/f2fs/super.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 033af907c3b1..7258a8bd6433 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -3364,6 +3364,15 @@ loff_t max_file_blocks(struct inode *inode)
 	leaf_count *= NIDS_PER_BLOCK;
 	result += leaf_count;
 
+	/* For compatibility with 4K crypto unit size, we must restrict
+	 * the max filesize to what can fit within U32_MAX 4K units.
+	 * Since the blocksize must currently be equal to the page size,
+	 * we can grab that from there. inode is NULL when setting up
+	 * the superblock.
+	 */
+
+	result = min(result, ((loff_t) U32_MAX * 4096) >> PAGE_SHIFT);
+
 	return result;
 }
 

base-commit: d346fa09abff46988de9267b67b6900d9913d5a2
-- 
2.43.0.rc2.451.g8631bc7472-goog

On Fri, Dec 01, 2023 at 07:03:47PM -0800, Daniel Rosenberg via Linux-f2fs-devel wrote:
> +	/* For compatibility with 4K crypto unit size, we must restrict
> +	 * the max filesize to what can fit within U32_MAX 4K units.
> +	 * Since the blocksize must currently be equal to the page size,
> +	 * we can grab that from there. inode is NULL when setting up
> +	 * the superblock.
> +	 */
> +
> +	result = min(result, ((loff_t) U32_MAX * 4096) >> PAGE_SHIFT);

This should be formatted like:

	/*
	 * For compatibility with ...
	 */
	result = ...

Also, the comment should mention that this is for the IV_INO_LBLK_64 and
IV_INO_LBLK_32 encryption settings.  There is no issue with the default
encryption setting.

Also, use F2FS_BLKSIZE_BITS instead of PAGE_SHIFT?

Also, the commit message could use:

	Fixes: d7e9a9037de2 ("f2fs: Support Block Size == Page Size")

- Eric