The compression filter driver is introduced as suggested by Max.
A sample usage of the filter can be found in the test #214.
Now, multiple clusters can be written compressed.
It is useful for the backup job.

v7:
  01: The 'zip_' prefix for the compression filter functions replaced with
    'compress_' one.
  02: .bdrv_co_preadv/pwritev (without _part) removed from the filter.
  03: .bdrv_refresh_limits amended.
  04: .bdrv_get_info added.
  05: In qapi/block-core.json, @compress: Since 5.0 was set.

  Discussed in the email thread with the message ID
  <1573488277-794975-1-git-send-email-andrey.shinkevich@virtuozzo.com>

Andrey Shinkevich (3):
  block: introduce compress filter driver
  qcow2: Allow writing compressed data of multiple clusters
  tests/qemu-iotests: add case to write compressed data of multiple
    clusters

 block/Makefile.objs        |   1 +
 block/filter-compress.c    | 201 +++++++++++++++++++++++++++++++++++++++++++++
 block/qcow2.c              | 102 +++++++++++++++++------
 qapi/block-core.json       |  10 ++-
 tests/qemu-iotests/214     |  43 ++++++++++
 tests/qemu-iotests/214.out |  14 ++++
 6 files changed, 340 insertions(+), 31 deletions(-)
 create mode 100644 block/filter-compress.c

-- 
1.8.3.1