Commit 734d8c15 authored by Josef Bacik's avatar Josef Bacik Committed by David Sterba

btrfs: add a comment describing block reserves

This is a giant comment at the top of block-rsv.c describing generally
how block reserves work.  It is purely about the block reserves
themselves, and nothing to do with how the actual reservation system
works.
Signed-off-by: default avatarJosef Bacik <josef@toxicpanda.com>
Signed-off-by: default avatarDavid Sterba <dsterba@suse.com>
parent 4cdfd930
......@@ -6,6 +6,98 @@
#include "space-info.h"
#include "transaction.h"
/*
* HOW DO BLOCK RESERVES WORK
*
* Think of block_rsv's as buckets for logically grouped metadata
* reservations. Each block_rsv has a ->size and a ->reserved. ->size is
* how large we want our block rsv to be, ->reserved is how much space is
* currently reserved for this block reserve.
*
* ->failfast exists for the truncate case, and is described below.
*
* NORMAL OPERATION
*
* -> Reserve
* Entrance: btrfs_block_rsv_add, btrfs_block_rsv_refill
*
* We call into btrfs_reserve_metadata_bytes() with our bytes, which is
* accounted for in space_info->bytes_may_use, and then add the bytes to
* ->reserved, and ->size in the case of btrfs_block_rsv_add.
*
* ->size is an over-estimation of how much we may use for a particular
* operation.
*
* -> Use
* Entrance: btrfs_use_block_rsv
*
* When we do a btrfs_alloc_tree_block() we call into btrfs_use_block_rsv()
* to determine the appropriate block_rsv to use, and then verify that
* ->reserved has enough space for our tree block allocation. Once
* successful we subtract fs_info->nodesize from ->reserved.
*
* -> Finish
* Entrance: btrfs_block_rsv_release
*
* We are finished with our operation, subtract our individual reservation
* from ->size, and then subtract ->size from ->reserved and free up the
* excess if there is any.
*
* There is some logic here to refill the delayed refs rsv or the global rsv
* as needed, otherwise the excess is subtracted from
* space_info->bytes_may_use.
*
* TYPES OF BLOCK RESERVES
*
* BLOCK_RSV_TRANS, BLOCK_RSV_DELOPS, BLOCK_RSV_CHUNK
* These behave normally, as described above, just within the confines of the
* lifetime of their particular operation (transaction for the whole trans
* handle lifetime, for example).
*
* BLOCK_RSV_GLOBAL
* It is impossible to properly account for all the space that may be required
* to make our extent tree updates. This block reserve acts as an overflow
* buffer in case our delayed refs reserve does not reserve enough space to
* update the extent tree.
*
* We can steal from this in some cases as well, notably on evict() or
* truncate() in order to help users recover from ENOSPC conditions.
*
* BLOCK_RSV_DELALLOC
* The individual item sizes are determined by the per-inode size
* calculations, which are described with the delalloc code. This is pretty
* straightforward, it's just the calculation of ->size encodes a lot of
* different items, and thus it gets used when updating inodes, inserting file
* extents, and inserting checksums.
*
* BLOCK_RSV_DELREFS
* We keep a running tally of how many delayed refs we have on the system.
* We assume each one of these delayed refs are going to use a full
* reservation. We use the transaction items and pre-reserve space for every
* operation, and use this reservation to refill any gap between ->size and
* ->reserved that may exist.
*
* From there it's straightforward, removing a delayed ref means we remove its
* count from ->size and free up reservations as necessary. Since this is
* the most dynamic block reserve in the system, we will try to refill this
* block reserve first with any excess returned by any other block reserve.
*
* BLOCK_RSV_EMPTY
* This is the fallback block reserve to make us try to reserve space if we
* don't have a specific bucket for this allocation. It is mostly used for
* updating the device tree and such, since that is a separate pool we're
* content to just reserve space from the space_info on demand.
*
* BLOCK_RSV_TEMP
* This is used by things like truncate and iput. We will temporarily
* allocate a block reserve, set it to some size, and then truncate bytes
* until we have no space left. With ->failfast set we'll simply return
* ENOSPC from btrfs_use_block_rsv() to signal that we need to unwind and try
* to make a new reservation. This is because these operations are
* unbounded, so we want to do as much work as we can, and then back off and
* re-reserve.
*/
static u64 block_rsv_release_bytes(struct btrfs_fs_info *fs_info,
struct btrfs_block_rsv *block_rsv,
struct btrfs_block_rsv *dest, u64 num_bytes,
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment