kernel/linux.git/fs/ecryptfs, branch master

Merge tag 'ecryptfs-7.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tyhicks/ecryptfs

2026-04-20T17:54:17+00:00

Pull eCryptfs updates from Tyler Hicks: - avoid unnecessary eCryptfs inode timestamp truncation by re-using the lower filesystem's time granularity - various small code cleanups - reorganize the setattr hook inode resizing to improve style and readability, remove an unnecessary memory allocation when shrinking, and to support an upcoming rework of the VFS interfaces involved in truncation * tag 'ecryptfs-7.1-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tyhicks/ecryptfs: ecryptfs: keep the lower iattr contained in truncate_upper ecryptfs: factor out a ecryptfs_iattr_to_lower helper ecryptfs: merge ecryptfs_inode_newsize_ok into truncate_upper ecryptfs: combine the two ATTR_SIZE blocks in ecryptfs_setattr ecryptfs: use ZERO_PAGE instead of allocating zeroed memory in truncate_upper ecryptfs: streamline truncate_upper ecryptfs: cleanup ecryptfs_setattr ecryptfs: Drop TODO comment in ecryptfs_derive_iv ecryptfs: Fix typo in ecryptfs_derive_iv function comment ecryptfs: Log function name only once in decode_and_decrypt_filename ecryptfs: Remove redundant if checks in encrypt_and_encode_filename ecryptfs: Fix tag number in encrypt_filename() error message ecryptfs: Use struct_size to improve process_response + send_miscdev ecryptfs: Replace memcpy + manual NUL termination with strscpy ecryptfs: Set s_time_gran to get correct time granularity

ecryptfs: keep the lower iattr contained in truncate_upper

2026-04-08T23:54:03+00:00

Currently the two callers of truncate_upper handle passing information very differently. ecryptfs_truncate passes a zeroed lower_ia and expects truncate_upper to fill it in from the upper ia created just for that, while ecryptfs_setattr passes a fully initialized lower_ia copied from the upper one. Both of them then call notify_change on the lower_ia. Switch to only passing the upper ia, and derive the lower ia from it inside truncate_upper, and call notify_change inside the function itself. Because the old name is misleading now, rename the resulting function to __ecryptfs_truncate as it deals with both the lower and upper inodes. Signed-off-by: Christoph Hellwig Signed-off-by: Tyler Hicks

ecryptfs: factor out a ecryptfs_iattr_to_lower helper

2026-04-08T23:54:03+00:00

Prepare for using the code to create a lower struct iattr in multiple places. Signed-off-by: Christoph Hellwig Signed-off-by: Tyler Hicks

ecryptfs: merge ecryptfs_inode_newsize_ok into truncate_upper

2026-04-08T23:54:03+00:00

Both callers of ecryptfs_inode_newsize_ok call truncate_upper right after. Merge ecryptfs_inode_newsize_ok into truncate_upper to simplify the logic. Signed-off-by: Christoph Hellwig Signed-off-by: Tyler Hicks

ecryptfs: combine the two ATTR_SIZE blocks in ecryptfs_setattr

2026-04-08T23:54:03+00:00

Simplify the logic in ecryptfs_setattr by combining the two ATTR_SIZE blocks. This initializes lower_ia before the size check, which is obviously correct as the size check doesn't look at it. Signed-off-by: Christoph Hellwig Signed-off-by: Tyler Hicks

ecryptfs: use ZERO_PAGE instead of allocating zeroed memory in truncate_upper

2026-04-08T23:54:03+00:00

Use the existing pre-zeroed memory instead of allocating a new chunk. Signed-off-by: Christoph Hellwig Signed-off-by: Tyler Hicks

ecryptfs: streamline truncate_upper

2026-04-08T23:52:51+00:00

Use a few strategic gotos to reduce indentation and keep the main flow outside of branches. Switch all touched comments to normal kernel style and avoid breaks in printed strings for all the code touched. Signed-off-by: Christoph Hellwig Signed-off-by: Tyler Hicks

ecryptfs: cleanup ecryptfs_setattr

2026-04-08T23:32:15+00:00

Initialize variables at declaration time where applicable and reformat conditionals to match the kernel coding style. Signed-off-by: Christoph Hellwig Signed-off-by: Tyler Hicks

ecryptfs: Drop TODO comment in ecryptfs_derive_iv

2026-04-02T22:27:52+00:00

Remove the TODO from 2006. eCryptfs is generally not receiving new features and changing the IV derivation is only likely to happen to address security concerns in the future. Signed-off-by: Thorsten Blum [tyhicks: Add the reasoning to the commit message] Signed-off-by: Tyler Hicks

ecryptfs: Fix typo in ecryptfs_derive_iv function comment

2026-04-02T22:22:02+00:00

s/vale/value/ Signed-off-by: Thorsten Blum Signed-off-by: Tyler Hicks