Orange Pi5 kernel

^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  1) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  2)  - support asynchronous operation -- add a per-fs 'reserved_space' count,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  3)    let each outstanding write reserve the _maximum_ amount of physical
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  4)    space it could take. Let GC flush the outstanding writes because the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  5)    reservations will necessarily be pessimistic. With this we could even
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  6)    do shared writable mmap, if we can have a fs hook for do_wp_page() to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  7)    make the reservation.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  8)  - disable compression in commit_write()?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300  9)  - fine-tune the allocation / GC thresholds
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10)  - chattr support - turning on/off and tuning compression per-inode
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11)  - checkpointing (do we need this? scan is quite fast)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12)  - make the scan code populate real inodes so read_inode just after 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) 	mount doesn't have to read the flash twice for large files.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) 	Make this a per-inode option, changeable with chattr, so you can
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) 	decide which inodes should be in-core immediately after mount.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16)  - test, test, test
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18)  - NAND flash support:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) 	- almost done :)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) 	- use bad block check instead of the hardwired byte check
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22)  - Optimisations:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23)    - Split writes so they go to two separate blocks rather than just c->nextblock.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) 	By writing _new_ nodes to one block, and garbage-collected REF_PRISTINE
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) 	nodes to a different one, we can separate clean nodes from those which
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) 	are likely to become dirty, and end up with blocks which are each far
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) 	closer to 100% or 0% clean, hence speeding up later GC progress dramatically.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28)    - Stop keeping name in-core with struct jffs2_full_dirent. If we keep the hash in 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29)      the full dirent, we only need to go to the flash in lookup() when we think we've
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30)      got a match, and in readdir(). 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31)    - Doubly-linked next_in_ino list to allow us to free obsoleted raw_node_refs immediately?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32)    - Remove size from jffs2_raw_node_frag. 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) 
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) dedekind:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) 1. __jffs2_flush_wbuf() has a strange 'pad' parameter. Eliminate.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) 2. get_sb()->build_fs()->scan() path... Why get_sb() removes scan()'s crap in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37)    case of failure? scan() does not clean everything. Fix.