public inbox for [email protected]
 help / color / mirror / Atom feed
From: Forza <[email protected]>
To: Stefan Metzmacher <[email protected]>, Jens Axboe <[email protected]>,
	[email protected]
Subject: Re: Stack trace with Samba VFS io_uring and large transfers
Date: Wed, 28 Jul 2021 11:39:02 +0200	[thread overview]
Message-ID: <[email protected]> (raw)
In-Reply-To: <[email protected]>



On 2021-07-28 02:16, Stefan Metzmacher wrote:
>> I am using Btrfs.
>>
>> My testing was done by exporting the share with
>>
>>    vfs objects = io_uring
>>    vfs objects = btrfs, io_uring
>>
>> Same results in both cases. Exporting with "vfs objects = btrfs" (no io_uring) works as expected.
> I don't think it makes a difference for the current problem, but I guess you want the following order instead:
> 
> vfs objects = io_uring, btrfs
> 
> metze

I haven't tested that combination. However the docs mention that 
io_uring must be last VFS object loaded.

https://www.samba.org/samba/docs/4.12/man-html/vfs_io_uring.8.html

"The io_uring VFS module enables asynchronous pread, pwrite and fsync 
using the io_uring infrastructure of Linux (>= 5.1). This provides much 
less overhead compared to the usage of the pthreadpool for async io.

This module SHOULD be listed last in any module stack as it requires 
real kernel file descriptors."

The manpage for vfs_btrfs mentions that btrfs is stackable and in their 
example they have btrfs loaded before vfs_shadow_copy.

https://www.samba.org/samba/docs/4.12/man-html/vfs_btrfs.8.html

What would be the implication of having io_uring before the btrfs module?

Regards
Forza

      parent reply	other threads:[~2021-07-28  9:39 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-24 15:46 Stack trace with Samba VFS io_uring and large transfers Forza
2021-07-24 17:04 ` Jens Axboe
2021-07-24 18:23   ` Forza
2021-07-24 19:44     ` Jens Axboe
2021-07-24 19:51       ` Forza
2021-07-26 16:07         ` Jens Axboe
2021-07-26 16:35           ` Jens Axboe
2021-07-28 10:02             ` Forza
2021-07-28 14:35               ` Jens Axboe
2021-07-28  0:16         ` Stefan Metzmacher
2021-07-28  2:34           ` Jens Axboe
2021-07-28  9:39           ` Forza [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    [email protected] \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox