On 8/13/21 3:32 AM, Dmitry Kadashev wrote:

On Fri, Jul 9, 2021 at 2:25 AM Jens Axboe [off-list ref] wrote:

quoted

On 7/8/21 12:34 PM, Linus Torvalds wrote:

quoted

On Wed, Jul 7, 2021 at 11:35 PM Dmitry Kadashev [off-list ref] wrote:

quoted

v9:
- reorder commits to keep io_uring ones nicely grouped at the end
- change 'fs:' to 'namei:' in related commit subjects, since this is
  what seems to be usually used in such cases

Ok, ack from me on this series, and as far as I'm concerned it can go
through the io_uring branch.

I'll queue it up in a separate branch. I'm assuming we're talking 5.15
at this point.

Is this going to be merged into 5.15? I'm still working on the follow-up
patch (well, right at this moment I'm actually on vacation, but will be
working on it when I'm back), but hopefully it does not have to be
merged in the same merge window / version? Especially given the fact
that Al prefers it to be a bigger refactoring of the ESTALE retries
rather than just moving bits and pieces to helper functions to simplify
the flow, see here:

https://lore.kernel.org/io-uring/20210715103600.3570667-1-dkadashev@gmail.com/ (local)

I added this to the for-5.15/io_uring-vfs branch:

https://git.kernel.dk/cgit/linux-block/log/?h=for-5.15/io_uring-vfs

had one namei.c conflict, set_nameidata() taking one more parameter, and
just a trivial conflict in each io_uring patch at the end. Can you double
check them?

-- 
Jens Axboe