diff options
| author | Linus Torvalds <torvalds@linux-foundation.org> | 2021-04-26 13:28:12 -0700 | 
|---|---|---|
| committer | Linus Torvalds <torvalds@linux-foundation.org> | 2021-04-26 13:28:12 -0700 | 
| commit | b5b3097d9cbb1eb3df0ade9507585e6e9e3b2385 (patch) | |
| tree | d9cec81907d2d7451e6a0b862ddfd79238c74437 /lib/mpi/mpi-cmp.c | |
| parent | befbfe07e646d9ffc5be1e2c943aefa5e23bf3b8 (diff) | |
| parent | 8e6c8fa9f2e95c88a642521a5da19a8e31748846 (diff) | |
Merge tag 'erofs-for-5.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs
Pull erofs updates from Gao Xiang:
 "In this cycle, we would like to introduce a new feature called big
  pcluster so EROFS can compress file data into more than 1 fs block and
  different pcluster size can be selected for each (sub-)files by
  design.
  The current EROFS test results on my laptop are [1]:
    Testscript: erofs-openbenchmark
    Testdata: enwik9 (1000000000 bytes)
     ________________________________________________________________
    |  file system  |   size    | seq read | rand read | rand9m read |
    |_______________|___________|_ MiB/s __|__ MiB/s __|___ MiB/s ___|
    |___erofs_4k____|_556879872_|_ 781.4 __|__ 55.3 ___|___ 25.3  ___|
    |___erofs_16k___|_452509696_|_ 864.8 __|_ 123.2 ___|___ 20.8  ___|
    |___erofs_32k___|_415223808_|_ 899.8 __|_ 105.8 _*_|___ 16.8 ____|
    |___erofs_64k___|_393814016_|_ 906.6 __|__ 66.6 _*_|___ 11.8 ____|
    |__squashfs_8k__|_556191744_|_  64.9 __|__ 19.3 ___|____ 9.1 ____|
    |__squashfs_16k_|_502661120_|_  98.9 __|__ 38.0 ___|____ 9.8 ____|
    |__squashfs_32k_|_458784768_|_ 115.4 __|__ 71.6 _*_|___ 10.0 ____|
    |_squashfs_128k_|_398204928_|_ 257.2 __|_ 253.8 _*_|___ 10.9 ____|
    |____ext4_4k____|____()_____|_ 786.6 __|__ 28.6 ___|___ 27.8 ____|
  which has been verified but I'd like warn it as experimental for a
  while. This matches erofs-utils dev branch and I'll also release a new
  userspace version for this later.
  Apart from that, several improvements are also included: eg complete a
  missing case for inplace I/O, optimize endio decompression logic for
  non-atomic contexts and support adjustable sliding window size, ... In
  addition to those, there are some cleanups as always.
  Summary:
   - avoid memory failure when applying rolling decompression
   - optimize endio decompression logic for non-atomic contexts
   - complete a missing case which can be safely selected for inplace
     I/O and thus decreasing more memory footprint
   - check unsupported on-disk inode i_format strictly
   - support adjustable lz4 sliding window size to decrease runtime
     memory footprint
   - support on-disk compression configurations
   - support big pcluster decompression
   - several code cleanups / spelling correction"
* tag 'erofs-for-5.13-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/xiang/erofs: (21 commits)
  erofs: enable big pcluster feature
  erofs: support decompress big pcluster for lz4 backend
  erofs: support parsing big pcluster compact indexes
  erofs: support parsing big pcluster compress indexes
  erofs: adjust per-CPU buffers according to max_pclusterblks
  erofs: add big physical cluster definition
  erofs: fix up inplace I/O pointer for big pcluster
  erofs: introduce physical cluster slab pools
  erofs: introduce multipage per-CPU buffers
  erofs: reserve physical_clusterbits[]
  erofs: Clean up spelling mistakes found in fs/erofs
  erofs: add on-disk compression configurations
  erofs: introduce on-disk lz4 fs configurations
  erofs: support adjust lz4 history window size
  erofs: introduce erofs_sb_has_xxx() helpers
  erofs: add unsupported inode i_format check
  erofs: don't use erofs_map_blocks() any more
  erofs: complete a missing case for inplace I/O
  erofs: use sync decompression for atomic contexts only
  erofs: use workqueue decompression for atomic contexts only
  ...
Diffstat (limited to 'lib/mpi/mpi-cmp.c')
0 files changed, 0 insertions, 0 deletions
