[GIT,PULL] overlayfs update for 6.7

Message ID 20231106065045.895874-1-amir73il@gmail.com
State New
Headers
Series [GIT,PULL] overlayfs update for 6.7 |

Pull-request

git://git.kernel.org/pub/scm/linux/kernel/git/overlayfs/vfs.git ovl-update-6.7

Message

Amir Goldstein Nov. 6, 2023, 6:50 a.m. UTC
  Hi Linus,

Please pull overlayfs update for 6.7.

Most of the changes in this update have been sitting in linux-next for
several weeks, based on two prerequisite vfs topic branches.
Last week, after the vfs branches were merged, I rebased overlayfs-next
to reduce noise.

Last week, I also added patches for new mount options to replace the
lowerdir append syntax that we disabled in v6.6-rc6 and v6.5.8.

This branch has gone through the usual overlayfs test routines and it
merges cleanly with master branch of the moment.

Thanks,
Amir.

----------------------------------------------------------------
The following changes since commit 14ab6d425e80674b6a0145f05719b11e82e64824:

  Merge tag 'vfs-6.7.ctime' of gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs (2023-10-30 09:47:13 -1000)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/overlayfs/vfs.git ovl-update-6.7

for you to fetch changes up to 24e16e385f2272b1a9df51337a5c32d28a29c7ad:

  ovl: add support for appending lowerdirs one by one (2023-10-31 00:13:02 +0200)

----------------------------------------------------------------
overlayfs update for 6.7

Contains the following patch sets:

- Overlayfs aio cleanups and fixes [1]

- Overlayfs lock ordering changes [2]

- Add support for nesting overlayfs private xattrs [3]

- Add new mount options for appending lowerdirs [4]

[1] https://lore.kernel.org/r/20230912173653.3317828-1-amir73il@gmail.com/
[2] https://lore.kernel.org/r/20230816152334.924960-1-amir73il@gmail.com/
[3] https://lore.kernel.org/r/cover.1694512044.git.alexl@redhat.com/
[4] https://lore.kernel.org/r/20231030120419.478228-1-amir73il@gmail.com/

----------------------------------------------------------------
Alexander Larsson (4):
      ovl: Add OVL_XATTR_TRUSTED/USER_PREFIX_LEN macros
      ovl: Support escaped overlay.* xattrs
      ovl: Add an alternative type of whiteout
      ovl: Add documentation on nesting of overlayfs mounts

Amir Goldstein (14):
      ovl: use simpler function to convert iocb to rw flags
      ovl: propagate IOCB_APPEND flag on writes to realfile
      ovl: punt write aio completion to workqueue
      ovl: protect copying of realinode attributes to ovl inode
      ovl: add helper ovl_file_modified()
      ovl: split ovl_want_write() into two helpers
      ovl: reorder ovl_want_write() after ovl_inode_lock()
      ovl: do not open/llseek lower file with upper sb_writers held
      ovl: do not encode lower fh with upper sb_writers held
      ovl: Move xattr support to new xattrs.c file
      ovl: remove unused code in lowerdir param parsing
      ovl: store and show the user provided lowerdir mount option
      ovl: refactor layer parsing helpers
      ovl: add support for appending lowerdirs one by one

 Documentation/filesystems/overlayfs.rst |  40 +++-
 fs/overlayfs/Makefile                   |   2 +-
 fs/overlayfs/copy_up.c                  | 142 +++++++++-----
 fs/overlayfs/dir.c                      |  64 +++----
 fs/overlayfs/export.c                   |   7 +-
 fs/overlayfs/file.c                     |  88 ++++++---
 fs/overlayfs/inode.c                    | 165 ++--------------
 fs/overlayfs/namei.c                    |  52 ++++--
 fs/overlayfs/overlayfs.h                |  72 +++++--
 fs/overlayfs/params.c                   | 322 +++++++++++++++++---------------
 fs/overlayfs/params.h                   |   1 +
 fs/overlayfs/readdir.c                  |  27 ++-
 fs/overlayfs/super.c                    |  92 ++-------
 fs/overlayfs/util.c                     | 115 +++++++++++-
 fs/overlayfs/xattrs.c                   | 271 +++++++++++++++++++++++++++
 fs/super.c                              |   1 +
 16 files changed, 929 insertions(+), 532 deletions(-)
 create mode 100644 fs/overlayfs/xattrs.c
  

Comments

Linus Torvalds Nov. 7, 2023, 7:54 p.m. UTC | #1
On Sun, 5 Nov 2023 at 22:50, Amir Goldstein <amir73il@gmail.com> wrote:
>
> - Overlayfs aio cleanups and fixes [1]
>
> - Overlayfs lock ordering changes [2]
>
> - Add support for nesting overlayfs private xattrs [3]
>
> - Add new mount options for appending lowerdirs [4]
>
> [1] https://lore.kernel.org/r/20230912173653.3317828-1-amir73il@gmail.com/
> [2] https://lore.kernel.org/r/20230816152334.924960-1-amir73il@gmail.com/
> [3] https://lore.kernel.org/r/cover.1694512044.git.alexl@redhat.com/
> [4] https://lore.kernel.org/r/20231030120419.478228-1-amir73il@gmail.com/

*Please* don't make me have to follow links just to see basic details.

Merge messages should stand on their own, not just point to "look,
here are the details in our lore archives". Yes, even when having
internet access is much more common, there are situations where it's
not there or is slow. Or maybe lore has issues. Or maybe people just
don't want to switch to a browser to look up details that may or may
not be relevant to them.

I copied the relevant stuff into my merge, but please don't make me do
this next time. Just give me a merge message with the details spelled
out, not some link to a cover letter for a patch series.

             Linus
  
pr-tracker-bot@kernel.org Nov. 7, 2023, 8:34 p.m. UTC | #2
The pull request you sent on Mon,  6 Nov 2023 08:50:45 +0200:

> git://git.kernel.org/pub/scm/linux/kernel/git/overlayfs/vfs.git ovl-update-6.7

has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/7f851936a0ca4b231224ee296cba28f9b1bc555e

Thank you!
  
Amir Goldstein Nov. 7, 2023, 9:07 p.m. UTC | #3
On Tue, Nov 7, 2023 at 9:54 PM Linus Torvalds
<torvalds@linux-foundation.org> wrote:
>
> On Sun, 5 Nov 2023 at 22:50, Amir Goldstein <amir73il@gmail.com> wrote:
> >
> > - Overlayfs aio cleanups and fixes [1]
> >
> > - Overlayfs lock ordering changes [2]
> >
> > - Add support for nesting overlayfs private xattrs [3]
> >
> > - Add new mount options for appending lowerdirs [4]
> >
> > [1] https://lore.kernel.org/r/20230912173653.3317828-1-amir73il@gmail.com/
> > [2] https://lore.kernel.org/r/20230816152334.924960-1-amir73il@gmail.com/
> > [3] https://lore.kernel.org/r/cover.1694512044.git.alexl@redhat.com/
> > [4] https://lore.kernel.org/r/20231030120419.478228-1-amir73il@gmail.com/
>
> *Please* don't make me have to follow links just to see basic details.
>

No problem.

> Merge messages should stand on their own, not just point to "look,
> here are the details in our lore archives". Yes, even when having
> internet access is much more common, there are situations where it's
> not there or is slow. Or maybe lore has issues. Or maybe people just
> don't want to switch to a browser to look up details that may or may
> not be relevant to them.
>
> I copied the relevant stuff into my merge, but please don't make me do
> this next time.

Thanks for adding this information - I certainly didn't intend for
you to do this extra work.

> Just give me a merge message with the details spelled
> out, not some link to a cover letter for a patch series.
>

OK.

To explain the reason why I add those links:

A while ago, I was helping out with backporting many xfs patches
to stable tree after a long period of neglection.

I found the need to point back from fix commits to patch series
because the cover letters often contained important information
relevant for backporting and testing.

So I hacked up a small b4 gadget to disassemble a PR
into lore links to patch series:
(*) you don't have to follow this link unless you are curious ;)
https://github.com/amir73il/b4/commit/f5966362a524182206d5c5e8a4f96fba5d4c92ca

At the time, I had wished that the links to the patch series
composing the PR would have been in the PR email
(doesn't really matter if you take them into the merge commit).

This is why I included this "debug info" of the PR.
Next time, if I include the lore links, I will separate
them to a different section, so it is clear that I do not
require them to be in the merge commit.

Thanks,
Amir.