Also seen with e2fsck; copied sully tlr dcp onto USB stick on the mac mini with 2.16.13, removed stick and put into laptop, e2fsck -f /dev/sdc1 and lots of errors...

Inode 69 is in use, but has dtime set.  Fix<y>? yes
Inode 69 has imagic flag set.  Clear<y>? yes
Inode 69 has illegal block(s).  Clear<y>? yes
Illegal block #0 (1853186645) in inode 69.  CLEARED.
Illegal block 0000001 (840988020) in inode 69.  CLEARED.
Illegal block 0000002 (875572785) in inode 69.  CLEARED.
Illegal block 0000003 (1684889888) in inode 69.  CLEARED.
Illegal block 0000004 (538981430) in inode 69.  CLEARED.
Illegal block 0000005 (538976288) in inode 69.  CLEARED.
Illegal block 0000006 (538976288) in inode 69.  CLEARED.
Illegal block 0000007 (538976288) in inode 69.  CLEARED.
Illegal block 0000011 (3455390720) in inode 69.  CLEARED.
Inode 69, i_size is 78128, should be 45056.  Fix<y>? yes
Inode 69, i_blocks is 538976288, should be 8.  Fix<y>? yes
Inode 71 is in use, but has dtime set.  Fix<y>? yes
Inode 71 has imagic flag set.  Clear<y>? yes
Inode 72 is in use, but has dtime set.  Fix<y>? yes
Inode 72 has imagic flag set.  Clear<y>? yes
Inode 73 is in use, but has dtime set.  Fix<y>? yes
Inode 73 has imagic flag set.  Clear<y>? yes
Inode 74 is in use, but has dtime set.  Fix<y>? yes
Inode 74 has imagic flag set.  Clear<y>? yes
Inode 75 is in use, but has dtime set.  Fix<y>? yes
Inode 75 has imagic flag set.  Clear<y>? yes
Inode 76 is in use, but has dtime set.  Fix<y>? yes
Inode 76 has imagic flag set.  Clear<y>? yes
Inode 77 is in use, but has dtime set.  Fix<y>? yes
Inode 77 has imagic flag set.  Clear<y>? yes
Inode 78 is in use, but has dtime set.  Fix<y>? yes
Inode 78 has imagic flag set.  Clear<y>? yes
Inode 79 is in use, but has dtime set.  Fix<y>? yes
Inode 79 has imagic flag set.  Clear<y>? yes
Inode 85 is in use, but has dtime set.  Fix<y>? yes
Inode 85 has imagic flag set.  Clear<y>? yes
Inode 87 is in use, but has dtime set.  Fix<y>? yes
Inode 87 has imagic flag set.  Clear<y>? yes
Special (device/socket/fifo) inode 87 has non-zero size.  Fix<y>? 
Inode 88 is in use, but has dtime set.  Fix<y>? yes
Inode 88 has imagic flag set.  Clear<y>? yes
Special (device/socket/fifo) inode 88 has non-zero size.  Fix<y>? yes
Inode 89 is in use, but has dtime set.  Fix<y>? yes
Inode 89 has imagic flag set.  Clear<y>? yes
Inode 89 has encrypt flag but no encryption extended attribute.
Clear flag<y>? yes
Inode 90 is in use, but has dtime set.  Fix<y>? yes
Inode 90 has imagic flag set.  Clear<y>? yes
Special (device/socket/fifo) inode 90 has non-zero size.  Fix<y>? yes
Inode 91 is in use, but has dtime set.  Fix<y>? yes
Inode 91 has imagic flag set.  Clear<y>? yes
Deleted inode 101 has zero dtime.  Fix<y>? yes
Inode 118 is in use, but has dtime set.  Fix<y>? yes
Inodes that were part of a corrupted orphan linked list found.  Fix<y>? yes
Inode 119 was part of the orphaned inode list.  FIXED.
Inode 121 is in use, but has dtime set.  Fix<y>? yes
Inode 122 was part of the orphaned inode list.  FIXED.
Inode 122 has INLINE_DATA_FL flag on filesystem without inline data support.
Clear<y>? yes
Inode 123 has EXTENTS_FL flag set on filesystem without extents support.
Clear<y>? yes
Inode 124 was part of the orphaned inode list.  FIXED.
Inode 124 has imagic flag set.  Clear<y>? yes
Inode 125 has EXTENTS_FL flag set on filesystem without extents support.
Clear<y>? yes
Inode 126 has the casefold flag set but is not a directory.  Clear flag<y>? yes
Inode 126 has EXTENTS_FL flag set on filesystem without extents support.
Clear<y>? yes
Inode 128 was part of the orphaned inode list.  FIXED.
Inode 133 has the casefold flag set but is not a directory.  Clear flag<y>? yes
Inode 133 has INLINE_DATA_FL flag on filesystem without inline data support.
Clear<y>? yes
Directory ??? has the casefold flag, but the
casefold feature is not enabled.  Clear flag<y>? yes
Inode 134 has INLINE_DATA_FL flag on filesystem without inline data support.
Clear<y>? yes
Inode 150 was part of the orphaned inode list.  FIXED.
Inode 151 has encrypt flag but no encryption extended attribute.
Clear flag<y>? yes
Deleted inode 152 has zero dtime.  Fix<y>? yes
Inode 153 was part of the orphaned inode list.  FIXED.
Inode 154 has encrypt flag but no encryption extended attribute.
Clear flag<y>? ``` etc. etc.

Doing diskutil eject after the write does not help.

Same problems doing the same thing from Linux.

Many errors are fixed by reverting 6d34064b3a4b07218b01f6cc7700ee2ce68a7542 in lwext4 which tried to use lazy setup of inode tables, a feature that seems only to exist in ext4.

Still some errors related to free block counts which seem to be reproducible in the unit tests if we copy the same DCP (Sully trailer). Trying to reproduce the same error with the xm trailer - if that passes I should add Sully to the private test set and change the unit test.

Believed fixed by 0774773de51df37d7af0d6ac920d304790a610c6 (reverting use of ext4-only lazy inode table initialisation) and ee865fa65f05e348cd4e0bce0552a2725ad5663a fixing incorrect block count in the last block group when the disk is not an exact number of blocks in length; at least, writing that same DCP to the same stick on mac and running e2fsck on Linux seems OK.