xfs: Use bulk page allocator for buffer cache
Hi Darrick,
Please pull the reviewed series from the signed tag detailed below. This has been updated with all the latested RVB tags and the couple of typos/whitespace issues you noticed when reviewing it. The branch is based on linux-xfs/master, and merges cleanly into the current for-next branch.
I sent a pull-req rather than posting the series again just for rvb updates as we already have enough noise on the list. Let me know if you would prefer patches over pull requests, or whether you want more information in the tags in future (e.g. series description for the merge commit) or anything else like that.
Cheers,
Dave.
The following changes since commit d07f6ca923ea0927a1024dfccafc5b53b61cfecc:
Linux 5.13-rc2 (2021-05-16 15:27:44 -0700)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs.git tags/xfs-buf-bulk-alloc-tag
for you to fetch changes up to 8bb870dee3c14ac0eded777a5c2d6d07a6cdd10c:
xfs: merge xfs_buf_allocate_memory (2021-06-01 13:40:37 +1000)
----------------------------------------------------------------
XFS buffer cache bulk page allocation
----------------------------------------------------------------
Christoph Hellwig (2):
xfs: simplify the b_page_count calculation
xfs: cleanup error handling in xfs_buf_get_map
Dave Chinner (8):
xfs: split up xfs_buf_allocate_memory
xfs: use xfs_buf_alloc_pages for uncached buffers
xfs: use alloc_pages_bulk_array() for buffers
xfs: merge _xfs_buf_get_pages()
xfs: move page freeing into _xfs_buf_free_pages()
xfs: remove ->b_offset handling for page backed buffers
xfs: get rid of xb_to_gfp()
xfs: merge xfs_buf_allocate_memory
fs/xfs/libxfs/xfs_ag.c | 1 - fs/xfs/xfs_buf.c | 305 +++++++++++++++++++++++++++++++++++++++++++++++------------------------------------------------------------------------- fs/xfs/xfs_buf.h | 3 +- 3 files changed, 120 insertions(+), 189 deletions(-)
Signature · 2 lines
Dave Chinner [email protected]
Quoted 24 lines · Dave Chinner
Hi Darrick,
Please pull the reviewed series from the signed tag detailed below. This has been updated with all the latested RVB tags and the couple of typos/whitespace issues you noticed when reviewing it. The branch is based on linux-xfs/master, and merges cleanly into the current for-next branch.
I sent a pull-req rather than posting the series again just for rvb updates as we already have enough noise on the list. Let me know if you would prefer patches over pull requests, or whether you want more information in the tags in future (e.g. series description for the merge commit) or anything else like that.
Cheers,
Dave.
The following changes since commit
d07f6ca923ea0927a1024dfccafc5b53b61cfecc:Linux 5.13-rc2 (2021-05-16 15:27:44 -0700)
are available in the Git repository at:
Quoted 9 lines
git://git.kernel.org/pub/scm/linux/kernel/git/dgc/linux-xfs.gittags/xfs-buf-bulk-alloc-tagfor you to fetch changes up to
8bb870dee3c14ac0eded777a5c2d6d07a6cdd10c:xfs: merge
xfs_buf_allocate_memory(2021-06-01 13:40:37 +1000)----------------------------------------------------------------
XFS buffer cache bulk page allocation
Would you mind pushing a new tag containing the cover letter for the series so that we can capture that in the git history?
--D
Quoted 23 lines
----------------------------------------------------------------
Christoph Hellwig (2):
xfs: simplify theb_page_countcalculation
xfs: cleanup error handling inxfs_buf_get_mapDave Chinner (8):
xfs: split upxfs_buf_allocate_memory
xfs: usexfs_buf_alloc_pagesfor uncached buffers
xfs: usealloc_pages_bulk_array()for buffers
xfs: merge_xfs_buf_get_pages()
xfs: move page freeing into_xfs_buf_free_pages()
xfs: remove ->b_offset handling for page backed buffers
xfs: get rid ofxb_to_gfp()
xfs: mergexfs_buf_allocate_memoryfs/xfs/libxfs/xfs_ag.c | 1 - fs/xfs/xfs_buf.c | 305 +++++++++++++++++++++++++++++++++++++++++++++++------------------------------------------------------------------------- fs/xfs/xfs_buf.h | 3 +- 3 files changed, 120 insertions(+), 189 deletions(-)Signature · 2 lines
Dave Chinner [email protected]