Skip to content

preparation for block layer simplification

Ming Lin linux-fsdevel linux-pm linux-raid lkml 1 version In review

Get series with b4
b4 am [email protected]

Versions newest first

  1. v3 Latest4 patches 9 repliesIn review

Changes in v3 from the cover letter

- rebase on top of Jen's bio-cnt branch
- drop patches 1 and 2 that were already upstream
- lib/iovec.c was gone, so move iov_count_pages to block/bio.c

Patches in v3 No applies recorded yet

#Subject and filesSizeRepliesApplied
1/4block: refactor iov_count_pages() from bio_{copy,map}_user_iov()block/bio.c+33 -371 reply
2/4md/raid10: make sync_request_write() call bio_copy_data()drivers/md/raid10.c+5 -151 reply
3/4fs: make _submit_bh consistent with generic bio chainingfs/buffer.c+2 -21 reply
4/4PM: submit bio in a sane way in cases without bio_chainkernel/power/block_io.c+18 -56 replies

Reviews trailers per patch

2 of 4 patches have a Reviewed-by or Acked-by.

PatchRafael J. WysockiChristoph HellwigPavel Machek
1/4
2/4R-b
3/4
4/4A-b T-b A-b

Trailers given in replies are filled; those carried in the patches from earlier versions are outlined with dashes. Scroll the table sideways for all 3 reviewers.

Read the v3 thread

Cover letter [PATCH v3 0/4] preparation for block layer simplification

I'd like to continue on Dongsu's effort to push Kent's block-stuff upstream.

This is a preparation series for simplifying block layer based on immutable biovecs, a spin off of the v2 of simplifying patchset. [1] The original goal of simplifying block layer was actually making generic_make_request() accept arbitrarily sized bios, and pushing the splitting down to the underlying drivers.

This patchset aims at cleaning up several parts that are independent of core changes in the block layer. Doing that, it could be possible to change block layer relatively easily without keeping up with many patches.

This patchset should be first applied prior to upcoming patchsets such as "simplify block layer based on immutable biovecs." This patchset itself should not bring any regression to end-users.

Changes in v3:
- rebase on top of Jen's bio-cnt branch
- drop patches 1 and 2 that were already upstream
- lib/iovec.c was gone, so move iov_count_pages to block/bio.c

Changes in v2:
- split up preparation patches from v1 into this separate series.
- In the patch 02, pass iov_iter by value to __bio_copy_iov(), and split into read/write variants, as suggested by Christoph Hellwig.
- minor changes like writing more commit messages etc.

[1] https://lkml.org/lkml/2014/12/22/128

Kent Overstreet (4):
block: refactor iov_count_pages() from bio_{copy,map}_user_iov() md/raid10: make sync_request_write() call bio_copy_data()
fs: make _submit_bh consistent with generic bio chaining
PM: submit bio in a sane way in cases without bio_chain

 block/bio.c             | 70 +++++++++++++++++++++++++++++++++-------------------------------------
 drivers/md/raid10.c     | 20 +++++---------------
 fs/buffer.c             |  4 ++--
 kernel/power/block_io.c | 23 ++++++++++++++++++-----
 4 files changed, 58 insertions(+), 59 deletions(-)