linux-f2fs-devel.lists.sourceforge.net archive mirror
 help / color / mirror / Atom feed
From: Chao Yu <chao@kernel.org>
To: Jaegeuk Kim <jaegeuk@kernel.org>, Chunhai Guo <guochunhai@vivo.com>
Cc: "linux-f2fs-devel@lists.sourceforge.net"
	<linux-f2fs-devel@lists.sourceforge.net>
Subject: Re: [f2fs-dev] [PATCH v2] f2fs: replace blk_finish_plug() with blk_flush_plug()
Date: Wed, 6 Sep 2023 21:31:24 +0800	[thread overview]
Message-ID: <aa4afdde-339d-1bdc-7ca3-1260aa7ae0bc@kernel.org> (raw)
In-Reply-To: <e0f7b14b-b50f-4cee-a187-35383147aa3e@vivo.com>

On 2023/9/5 21:42, Chunhai Guo wrote:
> 
> 
> On 2023/9/5 0:40, Jaegeuk Kim wrote:
>> On 09/04, Chao Yu wrote:
>>> On 2023/8/31 15:10, Chunhai Guo wrote:
>>>> The commit 344150999b7f ("f2fs: fix to avoid potential deadlock") only
>>>> requires unplugging current->plug. Using blk_finish_plug() is unnecessary
>>>> as it sets current->plug as NULL and prevents wb_writeback() from using
>>>> plug in subsequent loops. Instead, use blk_flush_plug() as a replacement.
>>>>
>>>> Signed-off-by: Chunhai Guo <guochunhai@vivo.com>
>>>> ---
>>>>     fs/f2fs/data.c | 3 +--
>>>>     fs/f2fs/node.c | 3 +--
>>>>     2 files changed, 2 insertions(+), 4 deletions(-)
>>>>
>>>> diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
>>>> index 916e317ac925..77b4a55d6d94 100644
>>>> --- a/fs/f2fs/data.c
>>>> +++ b/fs/f2fs/data.c
>>>> @@ -3346,8 +3346,7 @@ static int __f2fs_write_data_pages(struct address_space *mapping,
>>>>     		atomic_inc(&sbi->wb_sync_req[DATA]);
>>>>     	else if (atomic_read(&sbi->wb_sync_req[DATA])) {
>>>>     		/* to avoid potential deadlock */
>>>> -		if (current->plug)
>>>> -			blk_finish_plug(current->plug);
>>>
>>> What about?
>>>
>>> if (current->plug) {
>>> 	struct blk_plug *plug = current->plug;
>>>
>>> 	blk_finish_plug(plug);
>>> 	blk_start_plug(plug);
>>> }
>>>
>>> Jaegeuk, thoughts?
>>
>> By the way, do we really need to reuse current->plug again? After checkpoint
>> being done, we can use the plug in __f2fs_write_data_pages. Are there some
>> numbers to show the benefit?
>>
> 
> I don't have a number on this, but I think that reusing current->plug
> can help maintain the original intention of wb_writeback().
> Nevertheless, it can be a viable alternative to use the plug in
> __f2fs_write_data_pages().

I don't have numbers as well, but it can avoid break the intention of
per-writeback-work plug, which tries to merge IOs from all dirty inode.

Thanks,

> 
> Thanks.
>>>
>>> Thanks,
>>>
>>>> +		blk_flush_plug(current->plug, false);
>>>>     		goto skip_write;
>>>>     	}
>>>> diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
>>>> index ee2e1dd64f25..c4b5717a8c1a 100644
>>>> --- a/fs/f2fs/node.c
>>>> +++ b/fs/f2fs/node.c
>>>> @@ -2126,8 +2126,7 @@ static int f2fs_write_node_pages(struct address_space *mapping,
>>>>     		atomic_inc(&sbi->wb_sync_req[NODE]);
>>>>     	else if (atomic_read(&sbi->wb_sync_req[NODE])) {
>>>>     		/* to avoid potential deadlock */
>>>> -		if (current->plug)
>>>> -			blk_finish_plug(current->plug);
>>>> +		blk_flush_plug(current->plug, false);
>>>>     		goto skip_write;
>>>>     	}


_______________________________________________
Linux-f2fs-devel mailing list
Linux-f2fs-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel

      reply	other threads:[~2023-09-06 18:03 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-08-31  7:10 [f2fs-dev] [PATCH v2] f2fs: replace blk_finish_plug() with blk_flush_plug() Chunhai Guo via Linux-f2fs-devel
2023-08-31  7:43 ` Chao Yu
2023-09-04  8:21 ` Chao Yu
2023-09-04 16:40   ` Jaegeuk Kim
2023-09-05 13:42     ` Chunhai Guo via Linux-f2fs-devel
2023-09-06 13:31       ` Chao Yu [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 \
    --in-reply-to=aa4afdde-339d-1bdc-7ca3-1260aa7ae0bc@kernel.org \
    --to=chao@kernel.org \
    --cc=guochunhai@vivo.com \
    --cc=jaegeuk@kernel.org \
    --cc=linux-f2fs-devel@lists.sourceforge.net \
    /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;
as well as URLs for NNTP newsgroup(s).