lkml.org 
[lkml]   [2017]   [Nov]   [22]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Subject[PATCH 00/62] XArray November 2017 Edition
Date
From: Matthew Wilcox <mawilcox@microsoft.com>

I've lost count of the number of times I've posted the XArray before,
so time for a new numbering scheme. Here're two earlier versions,
https://lkml.org/lkml/2017/3/17/724
https://lwn.net/Articles/715948/ (this one's more loquacious in its
description of things that are better about the radix tree API than the
XArray).

This time around, I've gone for an approach of many small changes.
Unfortunately, that means you get 62 moderate patches instead of dozens
of big ones.

Some of these can and should go in independently of whether the XArray
is a good idea. The first four fix some things in the test-suite.
Patch 6 changes the API for the IDR cyclic allocation. Patch 7 changes
the API for the IDR 'extended' API that was recently added without my
review.

Patch 15 is messy and breaks any time anybody touches the page cache.
I'd like it to go in soon. All it's doing is removing the page cache's
private spinlock and having it explicitly use the spinlock now embedded
in the xarray.

Patches 16-30 are all the fun infrastructure, adding pieces of the
xarray API. This is probably a good place to focus review, particularly
if you're intent on critiquing the API.

Patches 31 onwards start actually using the new API, taking advantage of
the built-in locking and memory allocation to get rid of the preload
API. There's probably the most scope for bugs here; I've tried to
understand the locking schemes used in thirty different places, and
maybe I got some of them wrong.

There will need to be many more patches. Dozens, maybe hundreds. Once
we've converted all the calls to radix_tree_insert() over to
xa_store(), we can get rid of the GFP flags stored in the xarray head.
That'll make the mm people a little happier. Once we've got rid of
idr_preload(), we can get rid of the per-CPU cache of radix_tree_nodes.
I have a lot more enhancements on my todo list, but I'd like to get
something merged in the next cycle, and if I wait until my todo list is
empty, I will get nothing merged.

If you looked at earlier iterations, there was much more support for
huge pages. That's not in this revision; I couldn't debug the problems
I was seeing. I still have all that code, and I'm intending to put it
back in; it's just a bit down the todo list right now.

Matthew Wilcox (62):
tools: Make __test_and_clear_bit available
radix tree test suite: Remove ARRAY_SIZE
radix tree test suite: Check reclaim bit
idr test suite: Fix ida_test_random()
radix tree: Add a missing cast to gfp_t
idr: Make cursor explicit for cyclic allocation
idr: Rewrite extended IDR API
Explicitly include radix-tree.h
arm64: Turn flush_dcache_mmap_lock into a no-op
unicore32: Turn flush_dcache_mmap_lock into a no-op
Export __set_page_dirty
xfs: Rename xa_ elements to ail_
fscache: Use appropriate radix tree accessors
xarray: Add the xa_lock to the radix_tree_root
page cache: Use xa_lock
xarray: Replace exceptional entries
xarray: Change definition of sibling entries
xarray: Add definition of struct xarray
xarray: Define struct xa_node
xarray: Add xa_load
xarray: Add xa_get_tag, xa_set_tag and xa_clear_tag
xarray: Add xa_store
xarray: Add xa_cmpxchg
xarray: Add xa_for_each
xarray: Add xa_init
xarray: Add xas_for_each_tag
xarray: Add xa_get_entries and xa_get_tagged
xarray: Add xa_destroy
xarray: Add xas_prev_any
xarray: Add xas_find_any / xas_next_any
Convert IDR to use xarray
ida: Convert to using xarray
page cache: Convert page_cache_next_hole to XArray
page cache: Use xarray for adding pages
page cache: Convert page_cache_tree_delete to xarray
page cache: Convert find_get_entry to xarray
shmem: Convert replace to xarray
shmem: Convert shmem_confirm_swap to XArray
shmem: Convert find_swap_entry to XArray
shmem: Convert shmem_tag_pins to XArray
shmem: Convert shmem_wait_for_pins to XArray
vmalloc: Convert to xarray
brd: Convert to XArray
xfs: Convert m_perag_tree to XArray
xfs: Convert pag_ici_root to XArray
xfs: Convert xfs dquot to XArray
xfs: Convert mru cache to XArray
block: Remove IDR preloading
rxrpc: Remove IDR preloading
cgroup: Remove IDR wrappers
dca: Remove idr_preload calls
ipc: Remove call to idr_preload
irq: Remove call to idr_preload
scsi: Remove idr_preload in st driver
firewire: Remove call to idr_preload
drm: Remove drm_minor_lock and idr_preload
drm: Remove drm_syncobj_fd_to_handle
drm: Remove qxl driver IDR locks
drm: Replace virtio IDRs with IDAs
drm: Replace vmwgfx IDRs with IDAs
net: Redesign act_api use of IDR
mm: Convert page-writeback to XArray

Documentation/cgroup-v1/memory.txt | 2 +-
Documentation/vm/page_migration | 14 +-
arch/arm/include/asm/cacheflush.h | 6 +-
arch/arm64/include/asm/cacheflush.h | 6 +-
arch/nios2/include/asm/cacheflush.h | 6 +-
arch/parisc/include/asm/cacheflush.h | 6 +-
arch/powerpc/include/asm/book3s/64/pgtable.h | 4 +-
arch/powerpc/include/asm/nohash/64/pgtable.h | 4 +-
arch/powerpc/platforms/cell/spufs/sched.c | 2 +-
arch/unicore32/include/asm/cacheflush.h | 6 +-
block/genhd.c | 23 +-
drivers/block/brd.c | 87 +-
drivers/dca/dca-sysfs.c | 22 +-
drivers/firewire/core-cdev.c | 20 +-
drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c | 8 +-
drivers/gpu/drm/drm_dp_aux_dev.c | 5 +-
drivers/gpu/drm/drm_drv.c | 25 +-
drivers/gpu/drm/drm_gem.c | 27 +-
drivers/gpu/drm/drm_syncobj.c | 23 +-
drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c | 6 +-
drivers/gpu/drm/i915/i915_debugfs.c | 4 +-
drivers/gpu/drm/i915/i915_gem.c | 17 +-
drivers/gpu/drm/msm/msm_gem_submit.c | 10 +-
drivers/gpu/drm/qxl/qxl_cmd.c | 26 +-
drivers/gpu/drm/qxl/qxl_drv.h | 2 -
drivers/gpu/drm/qxl/qxl_kms.c | 2 -
drivers/gpu/drm/qxl/qxl_release.c | 12 +-
drivers/gpu/drm/vc4/vc4_gem.c | 4 +-
drivers/gpu/drm/virtio/virtgpu_drv.h | 6 +-
drivers/gpu/drm/virtio/virtgpu_kms.c | 18 +-
drivers/gpu/drm/virtio/virtgpu_vq.c | 12 +-
drivers/gpu/drm/vmwgfx/vmwgfx_drv.c | 6 +-
drivers/gpu/drm/vmwgfx/vmwgfx_drv.h | 2 +-
drivers/gpu/drm/vmwgfx/vmwgfx_resource.c | 28 +-
drivers/infiniband/core/cm.c | 5 +-
drivers/infiniband/hw/mlx4/cm.c | 3 +-
drivers/infiniband/hw/mlx4/mlx4_ib.h | 1 +
drivers/rapidio/rio_cm.c | 3 +-
drivers/rpmsg/qcom_glink_native.c | 7 +-
drivers/scsi/st.c | 18 +-
drivers/staging/lustre/lustre/llite/glimpse.c | 2 +-
drivers/staging/lustre/lustre/mdc/mdc_request.c | 10 +-
drivers/target/target_core_device.c | 4 +-
drivers/usb/host/xhci.h | 1 +
fs/afs/write.c | 2 +-
fs/btrfs/compression.c | 4 +-
fs/btrfs/extent_io.c | 8 +-
fs/btrfs/inode.c | 4 +-
fs/buffer.c | 13 +-
fs/cifs/file.c | 2 +-
fs/dax.c | 204 ++--
fs/f2fs/data.c | 6 +-
fs/f2fs/dir.c | 6 +-
fs/f2fs/gc.c | 2 +-
fs/f2fs/inline.c | 6 +-
fs/f2fs/node.c | 8 +-
fs/fs-writeback.c | 18 +-
fs/fscache/cookie.c | 2 +-
fs/fscache/object.c | 2 +-
fs/inode.c | 11 +-
fs/kernfs/dir.c | 5 +-
fs/nfsd/nfs4state.c | 4 +-
fs/nfsd/state.h | 1 +
fs/nilfs2/btnode.c | 20 +-
fs/nilfs2/page.c | 22 +-
fs/notify/inotify/inotify_user.c | 15 +-
fs/proc/loadavg.c | 2 +-
fs/proc/task_mmu.c | 2 +-
fs/xfs/libxfs/xfs_sb.c | 11 +-
fs/xfs/libxfs/xfs_sb.h | 2 +-
fs/xfs/xfs_aops.c | 15 +-
fs/xfs/xfs_buf_item.c | 10 +-
fs/xfs/xfs_dquot.c | 37 +-
fs/xfs/xfs_dquot_item.c | 11 +-
fs/xfs/xfs_icache.c | 142 +--
fs/xfs/xfs_icache.h | 10 +-
fs/xfs/xfs_inode.c | 24 +-
fs/xfs/xfs_inode_item.c | 22 +-
fs/xfs/xfs_log.c | 6 +-
fs/xfs/xfs_log_recover.c | 80 +-
fs/xfs/xfs_mount.c | 22 +-
fs/xfs/xfs_mount.h | 6 +-
fs/xfs/xfs_mru_cache.c | 71 +-
fs/xfs/xfs_qm.c | 32 +-
fs/xfs/xfs_qm.h | 18 +-
fs/xfs/xfs_trans.c | 18 +-
fs/xfs/xfs_trans_ail.c | 152 +--
fs/xfs/xfs_trans_buf.c | 4 +-
fs/xfs/xfs_trans_priv.h | 42 +-
include/drm/drm_file.h | 7 +-
include/linux/backing-dev.h | 12 +-
include/linux/fs.h | 68 +-
include/linux/fscache.h | 1 +
include/linux/fsnotify_backend.h | 1 +
include/linux/idr.h | 204 ++--
include/linux/kernfs.h | 1 +
include/linux/mm.h | 3 +-
include/linux/pagemap.h | 4 +-
include/linux/pid_namespace.h | 1 +
include/linux/radix-tree.h | 111 +-
include/linux/swapops.h | 19 +-
include/linux/xarray.h | 717 +++++++++++
include/net/act_api.h | 27 +-
include/net/sctp/sctp.h | 1 +
ipc/util.c | 28 +-
kernel/bpf/syscall.c | 8 +-
kernel/cgroup/cgroup.c | 63 +-
kernel/irq/timings.c | 4 +-
kernel/pid.c | 6 +-
kernel/pid_namespace.c | 12 +-
lib/Makefile | 2 +-
lib/dma-debug.c | 1 +
lib/idr.c | 470 ++++----
lib/radix-tree.c | 348 ++----
lib/xarray.c | 1444 +++++++++++++++++++++++
mm/filemap.c | 306 ++---
mm/huge_memory.c | 10 +-
mm/khugepaged.c | 51 +-
mm/madvise.c | 2 +-
mm/memcontrol.c | 4 +-
mm/migrate.c | 31 +-
mm/mincore.c | 2 +-
mm/page-writeback.c | 79 +-
mm/readahead.c | 4 +-
mm/rmap.c | 4 +-
mm/shmem.c | 196 ++-
mm/swap.c | 2 +-
mm/swap_state.c | 17 +-
mm/truncate.c | 34 +-
mm/vmalloc.c | 37 +-
mm/vmscan.c | 12 +-
mm/workingset.c | 34 +-
net/rxrpc/af_rxrpc.c | 2 +-
net/rxrpc/ar-internal.h | 1 +
net/rxrpc/conn_client.c | 21 +-
net/sched/act_api.c | 135 +--
net/sched/cls_basic.c | 20 +-
net/sched/cls_bpf.c | 20 +-
net/sched/cls_flower.c | 32 +-
net/sched/cls_u32.c | 44 +-
net/sctp/associola.c | 3 +-
net/sctp/protocol.c | 1 +
tools/include/asm-generic/bitops.h | 1 +
tools/include/asm-generic/bitops/atomic.h | 9 -
tools/include/asm-generic/bitops/non-atomic.h | 109 ++
tools/include/linux/spinlock.h | 2 +
tools/testing/radix-tree/.gitignore | 2 +
tools/testing/radix-tree/Makefile | 14 +-
tools/testing/radix-tree/idr-test.c | 28 +-
tools/testing/radix-tree/linux.c | 2 +-
tools/testing/radix-tree/linux/kernel.h | 2 -
tools/testing/radix-tree/linux/rcupdate.h | 2 +
tools/testing/radix-tree/linux/xarray.h | 2 +
tools/testing/radix-tree/multiorder.c | 53 +-
tools/testing/radix-tree/test.c | 8 +-
tools/testing/radix-tree/xarray-test.c | 122 ++
156 files changed, 4178 insertions(+), 2454 deletions(-)
create mode 100644 include/linux/xarray.h
create mode 100644 lib/xarray.c
create mode 100644 tools/include/asm-generic/bitops/non-atomic.h
create mode 100644 tools/testing/radix-tree/linux/xarray.h
create mode 100644 tools/testing/radix-tree/xarray-test.c

--
2.15.0

\
 
 \ /
  Last update: 2017-11-22 22:35    [W:0.207 / U:2.608 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site