[Qemu-devel] [PATCH v6 00/18] make dirty-bitmap byte-based

Eric Blake posted 18 patches 8 years, 1 month ago
Patches applied successfully (tree, apply log)
git fetch https://github.com/patchew-project/qemu tags/patchew/20170830210542.2153-1-eblake@redhat.com
Test FreeBSD passed
Test checkpatch passed
Test docker passed
Test s390x passed
There is a newer version of this series
include/block/block_int.h    |   2 +-
include/block/dirty-bitmap.h |  41 +++++---------
include/qemu/hbitmap.h       |   8 +--
block/io.c                   |   6 +-
block.c                      |   2 +-
block/backup.c               |   7 +--
block/dirty-bitmap.c         | 130 ++++++++++++++-----------------------------
block/mirror.c               |  76 +++++++++++--------------
block/qcow2-bitmap.c         |  57 +++++++++----------
block/qcow2.c                |  22 ++++----
migration/block.c            |  12 ++--
tests/test-hbitmap.c         |  10 ++--
util/hbitmap.c               |   8 +--
13 files changed, 154 insertions(+), 227 deletions(-)
[Qemu-devel] [PATCH v6 00/18] make dirty-bitmap byte-based
Posted by Eric Blake 8 years, 1 month ago
There are patches floating around to add NBD_CMD_BLOCK_STATUS,
but NBD wants to report status on byte granularity (even if the
reporting will probably be naturally aligned to sectors or even
much higher levels).  I've therefore started the task of
converting our block status code to report at a byte granularity
rather than sectors.

Now that 2.11 is open, I'm rebasing/reposting the remaining patches.

The overall conversion currently looks like:
part 1: bdrv_is_allocated (merged in 2.10, commit 51b0a488)
part 2: dirty-bitmap (this series, v5 was here [1])
part 3: bdrv_get_block_status (v3 is posted [2] and is mostly reviewed, but
needs a rebase)
part 4: .bdrv_co_block_status (v2 is posted [3], but needs a rebase)

Available as a tag at:
git fetch git://repo.or.cz/qemu/ericb.git nbd-byte-dirty-v6

[1] https://lists.gnu.org/archive/html/qemu-devel/2017-07/msg03512.html
[2] https://lists.gnu.org/archive/html/qemu-devel/2017-07/msg03853.html
[3] https://lists.gnu.org/archive/html/qemu-devel/2017-07/msg04370.html

Diff from v5:
- add another patch (more for ease of bookkeeping, as it was previously
posted independently)
- drop bug fixes that were hoisted into 2.10 (v5 1/18, plus 14/18)

001/18:[down] 'block: Make bdrv_img_create() size selection easier to read'
002/18:[----] [--] 'hbitmap: Rename serialization_granularity to serialization_align'
003/18:[----] [--] 'qcow2: Ensure bitmap serialization is aligned'
004/18:[----] [--] 'dirty-bitmap: Drop unused functions'
005/18:[----] [--] 'dirty-bitmap: Change bdrv_dirty_bitmap_size() to report bytes'
006/18:[----] [--] 'dirty-bitmap: Change bdrv_dirty_bitmap_*serialize*() to take bytes'
007/18:[----] [--] 'qcow2: Switch sectors_covered_by_bitmap_cluster() to byte-based'
008/18:[----] [--] 'dirty-bitmap: Set iterator start by offset, not sector'
009/18:[----] [--] 'dirty-bitmap: Change bdrv_dirty_iter_next() to report byte offset'
010/18:[----] [--] 'dirty-bitmap: Change bdrv_get_dirty_count() to report bytes'
011/18:[----] [--] 'dirty-bitmap: Change bdrv_get_dirty_locked() to take bytes'
012/18:[----] [--] 'dirty-bitmap: Change bdrv_[re]set_dirty_bitmap() to use bytes'
013/18:[----] [--] 'mirror: Switch mirror_dirty_init() to byte-based iteration'
014/18:[0004] [FC] 'qcow2: Switch qcow2_measure() to byte-based iteration'
015/18:[----] [--] 'qcow2: Switch load_bitmap_data() to byte-based iteration'
016/18:[----] [--] 'qcow2: Switch store_bitmap_data() to byte-based iteration'
017/18:[----] [--] 'dirty-bitmap: Switch bdrv_set_dirty() to bytes'
018/18:[----] [--] 'dirty-bitmap: Convert internal hbitmap size/granularity'

Eric Blake (18):
  block: Make bdrv_img_create() size selection easier to read
  hbitmap: Rename serialization_granularity to serialization_align
  qcow2: Ensure bitmap serialization is aligned
  dirty-bitmap: Drop unused functions
  dirty-bitmap: Change bdrv_dirty_bitmap_size() to report bytes
  dirty-bitmap: Change bdrv_dirty_bitmap_*serialize*() to take bytes
  qcow2: Switch sectors_covered_by_bitmap_cluster() to byte-based
  dirty-bitmap: Set iterator start by offset, not sector
  dirty-bitmap: Change bdrv_dirty_iter_next() to report byte offset
  dirty-bitmap: Change bdrv_get_dirty_count() to report bytes
  dirty-bitmap: Change bdrv_get_dirty_locked() to take bytes
  dirty-bitmap: Change bdrv_[re]set_dirty_bitmap() to use bytes
  mirror: Switch mirror_dirty_init() to byte-based iteration
  qcow2: Switch qcow2_measure() to byte-based iteration
  qcow2: Switch load_bitmap_data() to byte-based iteration
  qcow2: Switch store_bitmap_data() to byte-based iteration
  dirty-bitmap: Switch bdrv_set_dirty() to bytes
  dirty-bitmap: Convert internal hbitmap size/granularity

 include/block/block_int.h    |   2 +-
 include/block/dirty-bitmap.h |  41 +++++---------
 include/qemu/hbitmap.h       |   8 +--
 block/io.c                   |   6 +-
 block.c                      |   2 +-
 block/backup.c               |   7 +--
 block/dirty-bitmap.c         | 130 ++++++++++++++-----------------------------
 block/mirror.c               |  76 +++++++++++--------------
 block/qcow2-bitmap.c         |  57 +++++++++----------
 block/qcow2.c                |  22 ++++----
 migration/block.c            |  12 ++--
 tests/test-hbitmap.c         |  10 ++--
 util/hbitmap.c               |   8 +--
 13 files changed, 154 insertions(+), 227 deletions(-)

-- 
2.13.5


Re: [Qemu-devel] [PATCH v6 00/18] make dirty-bitmap byte-based
Posted by John Snow 8 years, 1 month ago

On 08/30/2017 05:05 PM, Eric Blake wrote:
> There are patches floating around to add NBD_CMD_BLOCK_STATUS,
> but NBD wants to report status on byte granularity (even if the
> reporting will probably be naturally aligned to sectors or even
> much higher levels).  I've therefore started the task of
> converting our block status code to report at a byte granularity
> rather than sectors.
> 
> Now that 2.11 is open, I'm rebasing/reposting the remaining patches.
> 
> The overall conversion currently looks like:
> part 1: bdrv_is_allocated (merged in 2.10, commit 51b0a488)
> part 2: dirty-bitmap (this series, v5 was here [1])
> part 3: bdrv_get_block_status (v3 is posted [2] and is mostly reviewed, but
> needs a rebase)
> part 4: .bdrv_co_block_status (v2 is posted [3], but needs a rebase)
> 
> Available as a tag at:
> git fetch git://repo.or.cz/qemu/ericb.git nbd-byte-dirty-v6
> 
> [1] https://lists.gnu.org/archive/html/qemu-devel/2017-07/msg03512.html
> [2] https://lists.gnu.org/archive/html/qemu-devel/2017-07/msg03853.html
> [3] https://lists.gnu.org/archive/html/qemu-devel/2017-07/msg04370.html
> 
> Diff from v5:
> - add another patch (more for ease of bookkeeping, as it was previously
> posted independently)
> - drop bug fixes that were hoisted into 2.10 (v5 1/18, plus 14/18)
> 
> 001/18:[down] 'block: Make bdrv_img_create() size selection easier to read'
> 002/18:[----] [--] 'hbitmap: Rename serialization_granularity to serialization_align'
> 003/18:[----] [--] 'qcow2: Ensure bitmap serialization is aligned'
> 004/18:[----] [--] 'dirty-bitmap: Drop unused functions'
> 005/18:[----] [--] 'dirty-bitmap: Change bdrv_dirty_bitmap_size() to report bytes'
> 006/18:[----] [--] 'dirty-bitmap: Change bdrv_dirty_bitmap_*serialize*() to take bytes'
> 007/18:[----] [--] 'qcow2: Switch sectors_covered_by_bitmap_cluster() to byte-based'
> 008/18:[----] [--] 'dirty-bitmap: Set iterator start by offset, not sector'
> 009/18:[----] [--] 'dirty-bitmap: Change bdrv_dirty_iter_next() to report byte offset'
> 010/18:[----] [--] 'dirty-bitmap: Change bdrv_get_dirty_count() to report bytes'
> 011/18:[----] [--] 'dirty-bitmap: Change bdrv_get_dirty_locked() to take bytes'
> 012/18:[----] [--] 'dirty-bitmap: Change bdrv_[re]set_dirty_bitmap() to use bytes'
> 013/18:[----] [--] 'mirror: Switch mirror_dirty_init() to byte-based iteration'
> 014/18:[0004] [FC] 'qcow2: Switch qcow2_measure() to byte-based iteration'
> 015/18:[----] [--] 'qcow2: Switch load_bitmap_data() to byte-based iteration'
> 016/18:[----] [--] 'qcow2: Switch store_bitmap_data() to byte-based iteration'
> 017/18:[----] [--] 'dirty-bitmap: Switch bdrv_set_dirty() to bytes'
> 018/18:[----] [--] 'dirty-bitmap: Convert internal hbitmap size/granularity'
> 
> Eric Blake (18):
>   block: Make bdrv_img_create() size selection easier to read
>   hbitmap: Rename serialization_granularity to serialization_align
>   qcow2: Ensure bitmap serialization is aligned
>   dirty-bitmap: Drop unused functions
>   dirty-bitmap: Change bdrv_dirty_bitmap_size() to report bytes
>   dirty-bitmap: Change bdrv_dirty_bitmap_*serialize*() to take bytes
>   qcow2: Switch sectors_covered_by_bitmap_cluster() to byte-based
>   dirty-bitmap: Set iterator start by offset, not sector
>   dirty-bitmap: Change bdrv_dirty_iter_next() to report byte offset
>   dirty-bitmap: Change bdrv_get_dirty_count() to report bytes
>   dirty-bitmap: Change bdrv_get_dirty_locked() to take bytes
>   dirty-bitmap: Change bdrv_[re]set_dirty_bitmap() to use bytes
>   mirror: Switch mirror_dirty_init() to byte-based iteration
>   qcow2: Switch qcow2_measure() to byte-based iteration
>   qcow2: Switch load_bitmap_data() to byte-based iteration
>   qcow2: Switch store_bitmap_data() to byte-based iteration
>   dirty-bitmap: Switch bdrv_set_dirty() to bytes
>   dirty-bitmap: Convert internal hbitmap size/granularity
> 
>  include/block/block_int.h    |   2 +-
>  include/block/dirty-bitmap.h |  41 +++++---------
>  include/qemu/hbitmap.h       |   8 +--
>  block/io.c                   |   6 +-
>  block.c                      |   2 +-
>  block/backup.c               |   7 +--
>  block/dirty-bitmap.c         | 130 ++++++++++++++-----------------------------
>  block/mirror.c               |  76 +++++++++++--------------
>  block/qcow2-bitmap.c         |  57 +++++++++----------
>  block/qcow2.c                |  22 ++++----
>  migration/block.c            |  12 ++--
>  tests/test-hbitmap.c         |  10 ++--
>  util/hbitmap.c               |   8 +--
>  13 files changed, 154 insertions(+), 227 deletions(-)
> 

Should this go through the bitmap tree, or since it's touching qcow2,
I'll let Kevin/Max/Stefan stage it?

--js