Linux-Fsdevel Archive on
help / color / mirror / Atom feed
From: Johannes Thumshirn <>
To: Christoph Hellwig <>
Cc: "" <>,
	Damien Le Moal <>,
	"" <>,
	Jens Axboe <>,
	"" <>
Subject: Re: [PATCH 2/2] zonefs: use zone-append for AIO as well
Date: Wed, 22 Jul 2020 15:00:49 +0000	[thread overview]
Message-ID: <> (raw)
In-Reply-To: <>

On 22/07/2020 16:52, Christoph Hellwig wrote:
> On Wed, Jul 22, 2020 at 12:43:21PM +0000, Johannes Thumshirn wrote:
>> On 21/07/2020 07:54, Christoph Hellwig wrote:
>>> On Mon, Jul 20, 2020 at 04:48:50PM +0000, Johannes Thumshirn wrote:
>>>> On 20/07/2020 15:45, Christoph Hellwig wrote:
>>>>> On Mon, Jul 20, 2020 at 10:21:18PM +0900, Johannes Thumshirn wrote:
>>>>>> On a successful completion, the position the data is written to is
>>>>>> returned via AIO's res2 field to the calling application.
>>>>> That is a major, and except for this changelog, undocumented ABI
>>>>> change.  We had the whole discussion about reporting append results
>>>>> in a few threads and the issues with that in io_uring.  So let's
>>>>> have that discussion there and don't mix it up with how zonefs
>>>>> writes data.  Without that a lot of the boilerplate code should
>>>>> also go away.
>>>> OK maybe I didn't remember correctly, but wasn't this all around 
>>>> io_uring and how we'd report the location back for raw block device
>>>> access?
>>> Report the write offset.  The author seems to be hell bent on making
>>> it block device specific, but that is a horrible idea as it is just
>>> as useful for normal file systems (or zonefs).
>> After having looked into io_uring I don't this there is anything that
>> prevents io_uring from picking up the write offset from ki_complete's
>> res2 argument. As of now io_uring ignores the filed but that can be 
>> changed.
> Sure.  Except for the fact that the io_uring CQE doesn't have space
> for it.  See the currently ongoing discussion on that..

That one I was aware of, but I thought once that discussion has settled
the write offset can be copied from res2 into what ever people have agreed
on by then.

>> So the only thing that needs to be done from a zonefs perspective is 
>> documenting the use of res2 and CC linux-aio and linux-abi (including
>> an update of the io_getevents man page).
>> Or am I completely off track now?
> Yes.  We should not have a different ABI just for zonefs.  We need to
> support this feature in a generic way and not as a weird one off for
> one filesystem and only with the legacy AIO interface.

OK, will have a look.

> Either way please make sure you properly separate the interface (
> using Write vs Zone Append in zonefs) from the interface (returning
> the actually written offset from appending writes), as they are quite
> separate issues.

So doing async RWF_APPEND writes using Zone Append isn't the problem here,
it's "only" the reporting of the write offset back to user-space? So once
we have sorted this out we can start issuing zone appends for zonefs async


  reply	other threads:[~2020-07-22 15:00 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-07-20 13:21 [PATCH 0/2] zonefs: use zone-append for aio with rwf append Johannes Thumshirn
2020-07-20 13:21 ` [PATCH 1/2] fs: fix kiocb ki_complete interface Johannes Thumshirn
2020-07-20 13:38   ` Christoph Hellwig
2020-07-20 13:43     ` Damien Le Moal
2020-07-20 13:47       ` Christoph Hellwig
2020-07-20 13:21 ` [PATCH 2/2] zonefs: use zone-append for AIO as well Johannes Thumshirn
2020-07-20 13:45   ` Christoph Hellwig
2020-07-20 16:48     ` Johannes Thumshirn
2020-07-21  5:54       ` Christoph Hellwig
2020-07-22 12:43         ` Johannes Thumshirn
2020-07-22 13:02           ` Damien Le Moal
2020-07-22 14:53             ` Christoph Hellwig
2020-07-22 14:51           ` Christoph Hellwig
2020-07-22 15:00             ` Johannes Thumshirn [this message]
2020-07-24 13:57             ` Kanchan Joshi
2020-07-27  3:12               ` Damien Le Moal
2020-07-21 12:43   ` Kanchan Joshi
2020-07-22 14:32     ` Johannes Thumshirn

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:

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

  git send-email \ \ \ \ \ \ \ \ \
    --subject='Re: [PATCH 2/2] zonefs: use zone-append for AIO as well' \

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).