https://bugzilla.kernel.org/show_bug.cgi?id=196405 --- Comment #4 from Andreas Dilger (adilger.kernelbugzilla@xxxxxxxxx) --- The limit is implemented as < 65000, which was somewhat below the hard limit of the ext4_inode.i_links_count __u16 field of 65535. There is no hard reason for 65000 except to allow some margin for detecting overflow, and reserved values if there was a need. I don't think it makes a big practical difference whether directory link counts are accurate up to 65000 or 65534 subdirectories, since very few systems have this many subdirectories. In theory, we could add an i_links_count_hi field to extend this to a 32-bit value, but it isn't clear if there is a big benefit from doing this? -- You are receiving this mail because: You are watching the assignee of the bug.
