[RFC PATCH v2 0/4] blk-cgroup: store blkcg in bio before blkcg_mutex conversion

Yu Kuai posted 4 patches 1 month, 2 weeks ago
There is a newer version of this series
Documentation/admin-guide/cgroup-v2.rst |   2 +-
block/bfq-cgroup.c                      |  16 +-
block/bfq-iosched.c                     |  19 +-
block/bio.c                             |  22 +-
block/blk-cgroup-fc-appid.c             |  10 +-
block/blk-cgroup.c                      | 348 ++++++++++++++----------
block/blk-cgroup.h                      |  82 ++++--
block/blk-core.c                        |   9 +-
block/blk-crypto-fallback.c             |   2 +-
block/blk-iocost.c                      |  12 +-
block/blk-iolatency.c                   |  11 +-
block/blk-ioprio.c                      |   2 +-
block/blk-throttle.c                    |   2 +-
block/blk-throttle.h                    |   2 +-
drivers/md/bcache/request.c             |   2 +-
drivers/md/dm.c                         |   2 +-
drivers/md/md.c                         |   2 +-
drivers/nvdimm/nd_virtio.c              |   2 +-
fs/gfs2/lops.c                          |   3 +-
include/linux/bio.h                     |  26 +-
include/linux/blk_types.h               |   9 +-
include/linux/blkdev.h                  |   2 +
include/linux/writeback.h               |   2 +-
mm/page_io.c                            |  13 +-
24 files changed, 357 insertions(+), 245 deletions(-)
[RFC PATCH v2 0/4] blk-cgroup: store blkcg in bio before blkcg_mutex conversion
Posted by Yu Kuai 1 month, 2 weeks ago
From: Yu Kuai <yukuai@fygo.io>

This RFC remains a preparatory series for the blkcg_mutex conversion
proposed in the related blkcg_mutex RFC v2 series [1].  That conversion
moves queue-local blkg topology synchronization from q->queue_lock to
q->blkcg_mutex, which is awkward while bios directly store queue-local
blkg references.

v1 made the stored bio association queue-independent by replacing bi_blkg
with bi_blkcg, but a bio-owned blkg reference still had to be recovered by
looking up the bio's blkcg and current request_queue.  Tao Cui reported that
cgroup removal deletes a dying blkg from the per-blkcg radix tree before a
throttled bio drops its reference.  A later lookup for that pinned blkg then
fails, triggers the warning in bio_pinned_blkg(), and leaks the reference.

v2 makes request_queue the authoritative lookup owner before converting the
bio association.  A queue-owned rhashtable, keyed by the blkcg CSS ID, keeps
dying blkgs discoverable until their references drain while q->blkg_list
remains available for ordered walks.  The bio conversion then stores and
pins the blkcg CSS, lazily creates a blkg only for users which need one, and
uses lookup-only access for completion and accounting paths which already
own a blkg reference.  Async bio punt state is finally moved from blkg to
blkcg so punting alone does not instantiate a queue-local blkg.

Changes since v1:

  - Add patch 1 to wait for every old blkg to leave q->blkg_list before a
    shared request_queue is rebound, instead of treating root_blkg == NULL
    as completion of asynchronous blkg teardown.
  - Add patch 2 to replace the per-blkcg radix tree and lookup hint with a
    request_queue rhashtable keyed by blkcg->css.id, as suggested by
    Christoph Hellwig.  Keep dying pinned blkgs in the hash until
    blkg_release() so bio-owned references remain discoverable.
  - Fold the v1 helper-only patch into patch 3, as suggested by Jan Kara and
    Christoph Hellwig, and make bio_blkcg() naturally return NULL for an
    unassociated bio.
  - Rework patch 3 so blkg_lookup_create() acquires the bio-owned reference,
    falls back to a live parent when creation or tryget fails, and updates
    bi_blkcg when the returned blkg belongs to an ancestor.
  - Make bio_blkg_lookup() lookup-only: it returns NULL unless BIO_BLKG_REF
    is already set.  Use bio_blkg() in the BFQ and IOCOST merge paths which
    may need to create a blkg, while keeping blk_cgroup_bio_start() and
    completion paths lookup-only.
  - Move CSS online-reference handling into
    bio_associate_blkcg_from_css(), including fallback to the root blkcg, so
    bio_associate_blkcg() does not take a redundant reference.
  - Keep the v1 async bio punt conversion as patch 4 and document that
    async_bio_lock protects async_bios.

Previous versions:
  v1: https://lore.kernel.org/r/20260804065313.2092022-1-yukuai@kernel.org

Related series:
  [1] RFC v2 blk-cgroup: protect blkgs with blkcg_mutex
      https://lore.kernel.org/r/20260724123037.3004560-1-yukuai@kernel.org

Yu Kuai (4):
  blk-cgroup: wait for old blkgs to leave queue before disk rebind
  blk-cgroup: use a request_queue rhashtable for blkg lookup
  blk-cgroup: store blkcg in bio instead of blkg
  blk-cgroup: move async bio punt state to blkcg

 Documentation/admin-guide/cgroup-v2.rst |   2 +-
 block/bfq-cgroup.c                      |  16 +-
 block/bfq-iosched.c                     |  19 +-
 block/bio.c                             |  22 +-
 block/blk-cgroup-fc-appid.c             |  10 +-
 block/blk-cgroup.c                      | 348 ++++++++++++++----------
 block/blk-cgroup.h                      |  82 ++++--
 block/blk-core.c                        |   9 +-
 block/blk-crypto-fallback.c             |   2 +-
 block/blk-iocost.c                      |  12 +-
 block/blk-iolatency.c                   |  11 +-
 block/blk-ioprio.c                      |   2 +-
 block/blk-throttle.c                    |   2 +-
 block/blk-throttle.h                    |   2 +-
 drivers/md/bcache/request.c             |   2 +-
 drivers/md/dm.c                         |   2 +-
 drivers/md/md.c                         |   2 +-
 drivers/nvdimm/nd_virtio.c              |   2 +-
 fs/gfs2/lops.c                          |   3 +-
 include/linux/bio.h                     |  26 +-
 include/linux/blk_types.h               |   9 +-
 include/linux/blkdev.h                  |   2 +
 include/linux/writeback.h               |   2 +-
 mm/page_io.c                            |  13 +-
 24 files changed, 357 insertions(+), 245 deletions(-)


base-commit: f2690679ecf3a3151688ebef766dd2512ff95854
-- 
2.51.0
Re: [RFC PATCH v2 0/4] blk-cgroup: store blkcg in bio before blkcg_mutex conversion
Posted by Christoph Hellwig 1 month, 2 weeks ago
What tree does this apply to?  I get failures both for Jens'
for-7.3/block and for-next trees in bfq-cgroups.c and bio.c

On Tue, Aug 11, 2026 at 02:47:40PM +0800, Yu Kuai wrote:
> From: Yu Kuai <yukuai@fygo.io>
> 
> This RFC remains a preparatory series for the blkcg_mutex conversion
> proposed in the related blkcg_mutex RFC v2 series [1].  That conversion
> moves queue-local blkg topology synchronization from q->queue_lock to
> q->blkcg_mutex, which is awkward while bios directly store queue-local
> blkg references.
> 
> v1 made the stored bio association queue-independent by replacing bi_blkg
> with bi_blkcg, but a bio-owned blkg reference still had to be recovered by
> looking up the bio's blkcg and current request_queue.  Tao Cui reported that
> cgroup removal deletes a dying blkg from the per-blkcg radix tree before a
> throttled bio drops its reference.  A later lookup for that pinned blkg then
> fails, triggers the warning in bio_pinned_blkg(), and leaks the reference.
> 
> v2 makes request_queue the authoritative lookup owner before converting the
> bio association.  A queue-owned rhashtable, keyed by the blkcg CSS ID, keeps
> dying blkgs discoverable until their references drain while q->blkg_list
> remains available for ordered walks.  The bio conversion then stores and
> pins the blkcg CSS, lazily creates a blkg only for users which need one, and
> uses lookup-only access for completion and accounting paths which already
> own a blkg reference.  Async bio punt state is finally moved from blkg to
> blkcg so punting alone does not instantiate a queue-local blkg.
> 
> Changes since v1:
> 
>   - Add patch 1 to wait for every old blkg to leave q->blkg_list before a
>     shared request_queue is rebound, instead of treating root_blkg == NULL
>     as completion of asynchronous blkg teardown.
>   - Add patch 2 to replace the per-blkcg radix tree and lookup hint with a
>     request_queue rhashtable keyed by blkcg->css.id, as suggested by
>     Christoph Hellwig.  Keep dying pinned blkgs in the hash until
>     blkg_release() so bio-owned references remain discoverable.
>   - Fold the v1 helper-only patch into patch 3, as suggested by Jan Kara and
>     Christoph Hellwig, and make bio_blkcg() naturally return NULL for an
>     unassociated bio.
>   - Rework patch 3 so blkg_lookup_create() acquires the bio-owned reference,
>     falls back to a live parent when creation or tryget fails, and updates
>     bi_blkcg when the returned blkg belongs to an ancestor.
>   - Make bio_blkg_lookup() lookup-only: it returns NULL unless BIO_BLKG_REF
>     is already set.  Use bio_blkg() in the BFQ and IOCOST merge paths which
>     may need to create a blkg, while keeping blk_cgroup_bio_start() and
>     completion paths lookup-only.
>   - Move CSS online-reference handling into
>     bio_associate_blkcg_from_css(), including fallback to the root blkcg, so
>     bio_associate_blkcg() does not take a redundant reference.
>   - Keep the v1 async bio punt conversion as patch 4 and document that
>     async_bio_lock protects async_bios.
> 
> Previous versions:
>   v1: https://lore.kernel.org/r/20260804065313.2092022-1-yukuai@kernel.org
> 
> Related series:
>   [1] RFC v2 blk-cgroup: protect blkgs with blkcg_mutex
>       https://lore.kernel.org/r/20260724123037.3004560-1-yukuai@kernel.org
> 
> Yu Kuai (4):
>   blk-cgroup: wait for old blkgs to leave queue before disk rebind
>   blk-cgroup: use a request_queue rhashtable for blkg lookup
>   blk-cgroup: store blkcg in bio instead of blkg
>   blk-cgroup: move async bio punt state to blkcg
> 
>  Documentation/admin-guide/cgroup-v2.rst |   2 +-
>  block/bfq-cgroup.c                      |  16 +-
>  block/bfq-iosched.c                     |  19 +-
>  block/bio.c                             |  22 +-
>  block/blk-cgroup-fc-appid.c             |  10 +-
>  block/blk-cgroup.c                      | 348 ++++++++++++++----------
>  block/blk-cgroup.h                      |  82 ++++--
>  block/blk-core.c                        |   9 +-
>  block/blk-crypto-fallback.c             |   2 +-
>  block/blk-iocost.c                      |  12 +-
>  block/blk-iolatency.c                   |  11 +-
>  block/blk-ioprio.c                      |   2 +-
>  block/blk-throttle.c                    |   2 +-
>  block/blk-throttle.h                    |   2 +-
>  drivers/md/bcache/request.c             |   2 +-
>  drivers/md/dm.c                         |   2 +-
>  drivers/md/md.c                         |   2 +-
>  drivers/nvdimm/nd_virtio.c              |   2 +-
>  fs/gfs2/lops.c                          |   3 +-
>  include/linux/bio.h                     |  26 +-
>  include/linux/blk_types.h               |   9 +-
>  include/linux/blkdev.h                  |   2 +
>  include/linux/writeback.h               |   2 +-
>  mm/page_io.c                            |  13 +-
>  24 files changed, 357 insertions(+), 245 deletions(-)
> 
> 
> base-commit: f2690679ecf3a3151688ebef766dd2512ff95854
> -- 
> 2.51.0
---end quoted text---
Re: [RFC PATCH v2 0/4] blk-cgroup: store blkcg in bio before blkcg_mutex conversion
Posted by yu kuai 1 month, 2 weeks ago
Hi,

在 2026/8/14 14:45, Christoph Hellwig 写道:
> What tree does this apply to?  I get failures both for Jens'
> for-7.3/block and for-next trees in bfq-cgroups.c and bio.c

Oh, just realize this set is cooked based on Jens' initial for-7.3/block
branch. With the other set that is already applied:

30d0c34b72f8 (HEAD -> block-7.3-blkg_list-rhashtable) blk-cgroup: move async bio punt state to blkcg
1aa13cdcc027 blk-cgroup: store blkcg in bio instead of blkg
058ff5f9c8ac blk-cgroup: use a request_queue rhashtable for blkg lookup
fefb6604e496 blk-cgroup: wait for old blkgs to leave queue before disk rebind
f2690679ecf3 (for-7.3/block) blk-cgroup: factor policy pd teardown loop into helper
ed3105cce735 blk-cgroup: skip dying blkg in blkcg_activate_policy()
415bd622f1a3 blk-cgroup: fix race between policy activation and blkg destruction
4cc0299aea5d blk-cgroup: protect q->blkg_list iteration in blkg_destroy_all() with blkcg_mutex
f5098b6bae76 (tag: v7.2-rc5, block-7.2) Linux 7.2-rc5

I'll rebase to the latest for-7.3/block branch in the next version.

>
> On Tue, Aug 11, 2026 at 02:47:40PM +0800, Yu Kuai wrote:
>> From: Yu Kuai <yukuai@fygo.io>
>>
>> This RFC remains a preparatory series for the blkcg_mutex conversion
>> proposed in the related blkcg_mutex RFC v2 series [1].  That conversion
>> moves queue-local blkg topology synchronization from q->queue_lock to
>> q->blkcg_mutex, which is awkward while bios directly store queue-local
>> blkg references.
>>
>> v1 made the stored bio association queue-independent by replacing bi_blkg
>> with bi_blkcg, but a bio-owned blkg reference still had to be recovered by
>> looking up the bio's blkcg and current request_queue.  Tao Cui reported that
>> cgroup removal deletes a dying blkg from the per-blkcg radix tree before a
>> throttled bio drops its reference.  A later lookup for that pinned blkg then
>> fails, triggers the warning in bio_pinned_blkg(), and leaks the reference.
>>
>> v2 makes request_queue the authoritative lookup owner before converting the
>> bio association.  A queue-owned rhashtable, keyed by the blkcg CSS ID, keeps
>> dying blkgs discoverable until their references drain while q->blkg_list
>> remains available for ordered walks.  The bio conversion then stores and
>> pins the blkcg CSS, lazily creates a blkg only for users which need one, and
>> uses lookup-only access for completion and accounting paths which already
>> own a blkg reference.  Async bio punt state is finally moved from blkg to
>> blkcg so punting alone does not instantiate a queue-local blkg.
>>
>> Changes since v1:
>>
>>    - Add patch 1 to wait for every old blkg to leave q->blkg_list before a
>>      shared request_queue is rebound, instead of treating root_blkg == NULL
>>      as completion of asynchronous blkg teardown.
>>    - Add patch 2 to replace the per-blkcg radix tree and lookup hint with a
>>      request_queue rhashtable keyed by blkcg->css.id, as suggested by
>>      Christoph Hellwig.  Keep dying pinned blkgs in the hash until
>>      blkg_release() so bio-owned references remain discoverable.
>>    - Fold the v1 helper-only patch into patch 3, as suggested by Jan Kara and
>>      Christoph Hellwig, and make bio_blkcg() naturally return NULL for an
>>      unassociated bio.
>>    - Rework patch 3 so blkg_lookup_create() acquires the bio-owned reference,
>>      falls back to a live parent when creation or tryget fails, and updates
>>      bi_blkcg when the returned blkg belongs to an ancestor.
>>    - Make bio_blkg_lookup() lookup-only: it returns NULL unless BIO_BLKG_REF
>>      is already set.  Use bio_blkg() in the BFQ and IOCOST merge paths which
>>      may need to create a blkg, while keeping blk_cgroup_bio_start() and
>>      completion paths lookup-only.
>>    - Move CSS online-reference handling into
>>      bio_associate_blkcg_from_css(), including fallback to the root blkcg, so
>>      bio_associate_blkcg() does not take a redundant reference.
>>    - Keep the v1 async bio punt conversion as patch 4 and document that
>>      async_bio_lock protects async_bios.
>>
>> Previous versions:
>>    v1: https://lore.kernel.org/r/20260804065313.2092022-1-yukuai@kernel.org
>>
>> Related series:
>>    [1] RFC v2 blk-cgroup: protect blkgs with blkcg_mutex
>>        https://lore.kernel.org/r/20260724123037.3004560-1-yukuai@kernel.org
>>
>> Yu Kuai (4):
>>    blk-cgroup: wait for old blkgs to leave queue before disk rebind
>>    blk-cgroup: use a request_queue rhashtable for blkg lookup
>>    blk-cgroup: store blkcg in bio instead of blkg
>>    blk-cgroup: move async bio punt state to blkcg
>>
>>   Documentation/admin-guide/cgroup-v2.rst |   2 +-
>>   block/bfq-cgroup.c                      |  16 +-
>>   block/bfq-iosched.c                     |  19 +-
>>   block/bio.c                             |  22 +-
>>   block/blk-cgroup-fc-appid.c             |  10 +-
>>   block/blk-cgroup.c                      | 348 ++++++++++++++----------
>>   block/blk-cgroup.h                      |  82 ++++--
>>   block/blk-core.c                        |   9 +-
>>   block/blk-crypto-fallback.c             |   2 +-
>>   block/blk-iocost.c                      |  12 +-
>>   block/blk-iolatency.c                   |  11 +-
>>   block/blk-ioprio.c                      |   2 +-
>>   block/blk-throttle.c                    |   2 +-
>>   block/blk-throttle.h                    |   2 +-
>>   drivers/md/bcache/request.c             |   2 +-
>>   drivers/md/dm.c                         |   2 +-
>>   drivers/md/md.c                         |   2 +-
>>   drivers/nvdimm/nd_virtio.c              |   2 +-
>>   fs/gfs2/lops.c                          |   3 +-
>>   include/linux/bio.h                     |  26 +-
>>   include/linux/blk_types.h               |   9 +-
>>   include/linux/blkdev.h                  |   2 +
>>   include/linux/writeback.h               |   2 +-
>>   mm/page_io.c                            |  13 +-
>>   24 files changed, 357 insertions(+), 245 deletions(-)
>>
>>
>> base-commit: f2690679ecf3a3151688ebef766dd2512ff95854
>> -- 
>> 2.51.0
> ---end quoted text---
>
-- 
Thanks,
Kuai
Re: [RFC PATCH v2 0/4] blk-cgroup: store blkcg in bio before blkcg_mutex conversion
Posted by Christoph Hellwig 1 month, 2 weeks ago
On Fri, Aug 14, 2026 at 03:53:44PM +0800, yu kuai wrote:
> I'll rebase to the latest for-7.3/block branch in the next version.

Thanks.  For anything complicated a link to the git branch can also
be helpful.