This removes the last ceph callers of find_or_create_page() and
find_get_page(), and one of the four remaining calls to
__page_cache_alloc().

The last patch converts the direct I/O put_bvecs() to folios as well.

The remaining inline data struct page use is the single-entry page array
in ceph_uninline_data(), due to the OSD client using page-based data
types.

Compile-tested only, as inline data support is deprecated.

---
Tal Zussman (6):
      ceph: use a folio in ceph_get_caps()
      ceph: use a folio in the inline read branch of ceph_filemap_fault()
      ceph: use a folio in the READ_INLINE path of ceph_read_iter()
      ceph: convert ceph_fill_inline_data() to take a folio
      ceph: convert r_locked_page to r_locked_folio
      ceph: use folios in put_bvecs()

 fs/ceph/addr.c       | 70 +++++++++++++++++++++++++---------------------------
 fs/ceph/caps.c       | 10 ++++----
 fs/ceph/file.c       | 32 +++++++++++++-----------
 fs/ceph/inode.c      | 14 +++++------
 fs/ceph/mds_client.h |  2 +-
 fs/ceph/super.h      | 10 ++++----
 6 files changed, 69 insertions(+), 69 deletions(-)
---
base-commit: 86f405187248b7a5a2104eba07f941d27593893e
change-id: 20260825-ceph-inline-folio-17bcdb200850

Best regards,
-- 
Tal Zussman <tz2294@columbia.edu>

Hi Tal,

Thanks for your patches. BTW, we are removing the inline code please
see https://lore.kernel.org/ceph-devel/20260729-b4-ceph-remove-inline-data-v2-1-db0c92dab8ee@clyso.com/.

BRs
Xiubo

Tal Zussman <tz2294@columbia.edu> 于2026年8月25日周二 14:49写道：
>
> This removes the last ceph callers of find_or_create_page() and
> find_get_page(), and one of the four remaining calls to
> __page_cache_alloc().
>
> The last patch converts the direct I/O put_bvecs() to folios as well.
>
> The remaining inline data struct page use is the single-entry page array
> in ceph_uninline_data(), due to the OSD client using page-based data
> types.
>
> Compile-tested only, as inline data support is deprecated.
>
> ---
> Tal Zussman (6):
>       ceph: use a folio in ceph_get_caps()
>       ceph: use a folio in the inline read branch of ceph_filemap_fault()
>       ceph: use a folio in the READ_INLINE path of ceph_read_iter()
>       ceph: convert ceph_fill_inline_data() to take a folio
>       ceph: convert r_locked_page to r_locked_folio
>       ceph: use folios in put_bvecs()
>
>  fs/ceph/addr.c       | 70 +++++++++++++++++++++++++---------------------------
>  fs/ceph/caps.c       | 10 ++++----
>  fs/ceph/file.c       | 32 +++++++++++++-----------
>  fs/ceph/inode.c      | 14 +++++------
>  fs/ceph/mds_client.h |  2 +-
>  fs/ceph/super.h      | 10 ++++----
>  6 files changed, 69 insertions(+), 69 deletions(-)
> ---
> base-commit: 86f405187248b7a5a2104eba07f941d27593893e
> change-id: 20260825-ceph-inline-folio-17bcdb200850
>
> Best regards,
> --
> Tal Zussman <tz2294@columbia.edu>
>
>

On 8/26/26 5:52 AM, Xiubo Li wrote:
> Hi Tal,
> 
> Thanks for your patches. BTW, we are removing the inline code please
> see https://lore.kernel.org/ceph-devel/20260729-b4-ceph-remove-inline-data-v2-1-db0c92dab8ee@clyso.com/.
> 
> BRs
> Xiubo
> 

Hi Xiubo,

Thanks, that's good to know.

I see that the patch you sent gates the inline code behind a config option
for a transitional period rather than removing it immediately.
In that case, I think it's still worth merging this series, as it removes
usage of some deprecated mm APIs that I'd like to remove in the next cycle.

Sashiko found some pre-existing issues with the inline data path, but if it's
being removed I'll leave them alone for now.

Thanks,
Tal