linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Nikolay Borisov <nborisov@suse.com>
To: Josef Bacik <josef@toxicpanda.com>,
	kernel-team@fb.com, linux-btrfs@vger.kernel.org
Subject: Re: [PATCH 7/8] btrfs: fix may_commit_transaction to deal with no partial filling
Date: Tue, 20 Aug 2019 11:20:47 +0300	[thread overview]
Message-ID: <c014e434-1451-79e7-52d6-32451893aa20@suse.com> (raw)
In-Reply-To: <20190816141952.19369-8-josef@toxicpanda.com>



On 16.08.19 г. 17:19 ч., Josef Bacik wrote:
> Now that we aren't partially filling tickets we may have some slack
> space left in the space_info.  We need to account for this in
> may_commit_transaction, otherwise we may choose to not commit the
> transaction despite it actually having enough space to satisfy our
> ticket.
> 
> Calculate the free space we have in the space_info, if any.  Then check
> to see if its >= the amount of bytes_needed after we've accounted for
> the space being used by the delayed refs rsv.  If it's not subtract it
> from the bytes_needed before doing the final pinned check.  If we still
> don't have enough space then we are truly out of space.
> 
> Signed-off-by: Josef Bacik <josef@toxicpanda.com>
> ---
>  fs/btrfs/space-info.c | 19 +++++++++++++++++++
>  1 file changed, 19 insertions(+)
> 
> diff --git a/fs/btrfs/space-info.c b/fs/btrfs/space-info.c
> index bd485be783b8..f79afdc04925 100644
> --- a/fs/btrfs/space-info.c
> +++ b/fs/btrfs/space-info.c
> @@ -471,12 +471,19 @@ static int may_commit_transaction(struct btrfs_fs_info *fs_info,
>  	struct btrfs_trans_handle *trans;
>  	u64 bytes_needed;
>  	u64 reclaim_bytes = 0;
> +	u64 cur_free_bytes = 0;
>  
>  	trans = (struct btrfs_trans_handle *)current->journal_info;
>  	if (trans)
>  		return -EAGAIN;
>  
>  	spin_lock(&space_info->lock);
> +	cur_free_bytes = btrfs_space_info_used(space_info, true);
> +	if (cur_free_bytes < space_info->total_bytes)
> +		cur_free_bytes = space_info->total_bytes - cur_free_bytes;
> +	else
> +		cur_free_bytes = 0;
> +

Why don't you subtract  cur_free_bytes from bytes_needed right here,
giving you more chance to hit the:

if (reclaim_bytes >= bytes_needed)
                goto commit;

>  	if (!list_empty(&space_info->priority_tickets))
>  		ticket = list_first_entry(&space_info->priority_tickets,
>  					  struct reserve_ticket, list);
> @@ -522,6 +529,18 @@ static int may_commit_transaction(struct btrfs_fs_info *fs_info,
>  		goto commit;
>  	bytes_needed -= reclaim_bytes;
>  
> +	/*
> +	 * We don't partially fill tickets anymore, so we could have some free
> +	 * bytes in the space_info already, just not enough to satisfy the
> +	 * ticket.  If bytes_needed is already below cur_free_bytes after taking
> +	 * away the delayed refs and delayed rsv's then we can commit.
> +	 * Otherwise subtract our cur_free_bytes from bytes_needed before we
> +	 * check pinned.
> +	 */
> +	if (bytes_needed <= cur_free_bytes)
> +		goto commit;

If you do that you won't need that check because you have already
accounted for the current free space.

> +	bytes_needed -= cur_free_bytes;

And this line will be gone as well.

> +
>  	if (__percpu_counter_compare(&space_info->total_bytes_pinned,
>  				   bytes_needed,
>  				   BTRFS_TOTAL_BYTES_PINNED_BATCH) < 0)
> 

  reply	other threads:[~2019-08-20  8:20 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-16 14:19 [PATCH 0/8][v2] Rework reserve ticket handling Josef Bacik
2019-08-16 14:19 ` [PATCH 1/8] btrfs: do not allow reservations if we have pending tickets Josef Bacik
2019-08-19 12:54   ` Nikolay Borisov
2019-08-19 12:57     ` Josef Bacik
2019-08-16 14:19 ` [PATCH 2/8] btrfs: roll tracepoint into btrfs_space_info_update helper Josef Bacik
2019-08-16 14:19 ` [PATCH 3/8] btrfs: add space reservation tracepoint for reserved bytes Josef Bacik
2019-08-16 14:19 ` [PATCH 4/8] btrfs: rework btrfs_space_info_add_old_bytes Josef Bacik
2019-08-16 14:19 ` [PATCH 5/8] btrfs: refactor the ticket wakeup code Josef Bacik
2019-08-19 13:58   ` Nikolay Borisov
2019-08-16 14:19 ` [PATCH 6/8] btrfs: rework wake_all_tickets Josef Bacik
2019-08-19 14:49   ` Nikolay Borisov
2019-08-19 15:06     ` Josef Bacik
2019-08-20  7:51       ` Nikolay Borisov
2019-08-16 14:19 ` [PATCH 7/8] btrfs: fix may_commit_transaction to deal with no partial filling Josef Bacik
2019-08-20  8:20   ` Nikolay Borisov [this message]
2019-08-16 14:19 ` [PATCH 8/8] btrfs: remove orig_bytes from reserve_ticket Josef Bacik
2019-08-20  8:28   ` Nikolay Borisov
2019-08-21 14:07     ` David Sterba

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=c014e434-1451-79e7-52d6-32451893aa20@suse.com \
    --to=nborisov@suse.com \
    --cc=josef@toxicpanda.com \
    --cc=kernel-team@fb.com \
    --cc=linux-btrfs@vger.kernel.org \
    /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).