Linux kernel (f2fs) vulnerabilities

A flaw was found in the Linux kernel F2FS filesystem where extending a file size across an unaligned EOF boundary did not zero out post-EOF data in the partial page. As a result, stale disk data beyond the previous EOF could be exposed after remounting or crash recovery.

In the Linux kernel, __allocate_data_block() in the f2fs filesystem increments the valid block count via inc_valid_block_count() when allocating a new data block, but if the subsequent f2fs_allocate_data_block() fails, the function returns the error without rolling back that increment. This causes a permanent valid block count leak.

In the Linux kernel, f2fs_sanity_check_node_footer() is not passed the correct folio->index value, causing it to check the same nid incorrectly. This can lead to improper node consistency verification in the f2fs filesystem.

In the Linux kernel, the f2fs filesystem lacked a sanity check on the node block address (blkaddr) in truncate_node(). On a crafted filesystem image, a corrupted NAT entry can cause a kernel panic during block invalidation.