public inbox for linux-xfs@vger.kernel.org
 help / color / mirror / Atom feed
From: "Darrick J. Wong" <djwong@kernel.org>
To: Christoph Hellwig <hch@lst.de>
Cc: Chandan Babu R <chandan.babu@oracle.com>,
	Dave Chinner <david@fromorbit.com>,
	linux-xfs@vger.kernel.org
Subject: Re: [PATCH 11/13] xfs: rework splitting of indirect block reservations
Date: Wed, 27 Mar 2024 08:14:56 -0700	[thread overview]
Message-ID: <20240327151456.GZ6390@frogsfrogsfrogs> (raw)
In-Reply-To: <20240327110318.2776850-12-hch@lst.de>

On Wed, Mar 27, 2024 at 12:03:16PM +0100, Christoph Hellwig wrote:
> Move the check if we have enough indirect blocks and the stealing of
> the deleted extent blocks out of xfs_bmap_split_indlen and into the
> caller to prepare for handling delayed allocation of RT extents that
> can't easily be stolen.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Looks correct now,
Reviewed-by: Darrick J. Wong <djwong@kernel.org>

--D

> ---
>  fs/xfs/libxfs/xfs_bmap.c | 38 ++++++++++++++++----------------------
>  1 file changed, 16 insertions(+), 22 deletions(-)
> 
> diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
> index 131d4f063b660a..9d0b7caa9a036c 100644
> --- a/fs/xfs/libxfs/xfs_bmap.c
> +++ b/fs/xfs/libxfs/xfs_bmap.c
> @@ -4829,31 +4829,17 @@ xfs_bmapi_remap(
>   * ores == 1). The number of stolen blocks is returned. The availability and
>   * subsequent accounting of stolen blocks is the responsibility of the caller.
>   */
> -static xfs_filblks_t
> +static void
>  xfs_bmap_split_indlen(
>  	xfs_filblks_t			ores,		/* original res. */
>  	xfs_filblks_t			*indlen1,	/* ext1 worst indlen */
> -	xfs_filblks_t			*indlen2,	/* ext2 worst indlen */
> -	xfs_filblks_t			avail)		/* stealable blocks */
> +	xfs_filblks_t			*indlen2)	/* ext2 worst indlen */
>  {
>  	xfs_filblks_t			len1 = *indlen1;
>  	xfs_filblks_t			len2 = *indlen2;
>  	xfs_filblks_t			nres = len1 + len2; /* new total res. */
> -	xfs_filblks_t			stolen = 0;
>  	xfs_filblks_t			resfactor;
>  
> -	/*
> -	 * Steal as many blocks as we can to try and satisfy the worst case
> -	 * indlen for both new extents.
> -	 */
> -	if (ores < nres && avail)
> -		stolen = XFS_FILBLKS_MIN(nres - ores, avail);
> -	ores += stolen;
> -
> -	 /* nothing else to do if we've satisfied the new reservation */
> -	if (ores >= nres)
> -		return stolen;
> -
>  	/*
>  	 * We can't meet the total required reservation for the two extents.
>  	 * Calculate the percent of the overall shortage between both extents
> @@ -4898,8 +4884,6 @@ xfs_bmap_split_indlen(
>  
>  	*indlen1 = len1;
>  	*indlen2 = len2;
> -
> -	return stolen;
>  }
>  
>  int
> @@ -4915,7 +4899,7 @@ xfs_bmap_del_extent_delay(
>  	struct xfs_bmbt_irec	new;
>  	int64_t			da_old, da_new, da_diff = 0;
>  	xfs_fileoff_t		del_endoff, got_endoff;
> -	xfs_filblks_t		got_indlen, new_indlen, stolen;
> +	xfs_filblks_t		got_indlen, new_indlen, stolen = 0;
>  	uint32_t		state = xfs_bmap_fork_to_state(whichfork);
>  	uint64_t		fdblocks;
>  	int			error = 0;
> @@ -4994,8 +4978,19 @@ xfs_bmap_del_extent_delay(
>  		new_indlen = xfs_bmap_worst_indlen(ip, new.br_blockcount);
>  
>  		WARN_ON_ONCE(!got_indlen || !new_indlen);
> -		stolen = xfs_bmap_split_indlen(da_old, &got_indlen, &new_indlen,
> -						       del->br_blockcount);
> +		/*
> +		 * Steal as many blocks as we can to try and satisfy the worst
> +		 * case indlen for both new extents.
> +		 */
> +		da_new = got_indlen + new_indlen;
> +		if (da_new > da_old) {
> +			stolen = XFS_FILBLKS_MIN(da_new - da_old,
> +						 del->br_blockcount);
> +			da_old += stolen;
> +		}
> +		if (da_new > da_old)
> +			xfs_bmap_split_indlen(da_old, &got_indlen, &new_indlen);
> +		da_new = got_indlen + new_indlen;
>  
>  		got->br_startblock = nullstartblock((int)got_indlen);
>  
> @@ -5007,7 +5002,6 @@ xfs_bmap_del_extent_delay(
>  		xfs_iext_next(ifp, icur);
>  		xfs_iext_insert(ip, icur, &new, state);
>  
> -		da_new = got_indlen + new_indlen - stolen;
>  		del->br_blockcount -= stolen;
>  		break;
>  	}
> -- 
> 2.39.2
> 
> 

  reply	other threads:[~2024-03-27 15:14 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-03-27 11:03 bring back RT delalloc support v5 Christoph Hellwig
2024-03-27 11:03 ` [PATCH 01/13] xfs: make XFS_TRANS_LOWMODE match the other XFS_TRANS_ definitions Christoph Hellwig
2024-03-28  3:08   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 02/13] xfs: refactor realtime inode locking Christoph Hellwig
2024-03-28  3:15   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 03/13] xfs: free RT extents after updating the bmap btree Christoph Hellwig
2024-03-27 14:55   ` Darrick J. Wong
2024-03-27 17:03     ` Christoph Hellwig
2024-03-28  4:13   ` Dave Chinner
2024-03-29  4:14     ` Christoph Hellwig
2024-03-30 21:55       ` Dave Chinner
2024-03-27 11:03 ` [PATCH 04/13] xfs: move RT inode locking out of __xfs_bunmapi Christoph Hellwig
2024-03-27 15:07   ` Darrick J. Wong
2024-03-27 17:06     ` Christoph Hellwig
2024-03-27 18:06       ` Darrick J. Wong
2024-03-27 18:12         ` Christoph Hellwig
2024-03-28  4:15   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 05/13] xfs: block deltas in xfs_trans_unreserve_and_mod_sb must be positive Christoph Hellwig
2024-03-28  4:16   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 06/13] xfs: split xfs_mod_freecounter Christoph Hellwig
2024-03-27 15:12   ` Darrick J. Wong
2024-03-28  4:18   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 07/13] xfs: reinstate RT support in xfs_bmapi_reserve_delalloc Christoph Hellwig
2024-03-28  4:20   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 08/13] xfs: cleanup fdblock/frextent accounting in xfs_bmap_del_extent_delay Christoph Hellwig
2024-03-28  4:22   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 09/13] xfs: support RT inodes in xfs_mod_delalloc Christoph Hellwig
2024-03-27 15:20   ` Darrick J. Wong
2024-03-27 17:05     ` Christoph Hellwig
2024-03-28  4:27   ` Dave Chinner
2024-03-28  4:34     ` Christoph Hellwig
2024-03-28  4:42       ` Dave Chinner
2024-03-28  8:25         ` Christoph Hellwig
2024-03-28 21:25           ` Dave Chinner
2024-03-27 11:03 ` [PATCH 10/13] xfs: look at m_frextents in xfs_iomap_prealloc_size for RT allocations Christoph Hellwig
2024-03-28  4:32   ` Dave Chinner
2024-03-28  4:34     ` Christoph Hellwig
2024-03-27 11:03 ` [PATCH 11/13] xfs: rework splitting of indirect block reservations Christoph Hellwig
2024-03-27 15:14   ` Darrick J. Wong [this message]
2024-03-28  4:35   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 12/13] xfs: stop the steal (of data blocks for RT indirect blocks) Christoph Hellwig
2024-03-27 15:18   ` Darrick J. Wong
2024-03-28  4:36   ` Dave Chinner
2024-03-27 11:03 ` [PATCH 13/13] xfs: reinstate delalloc for RT inodes (if sb_rextsize == 1) Christoph Hellwig
2024-03-28  4:39   ` Dave Chinner
  -- strict thread matches above, loose matches on Subject: below --
2024-04-22 11:20 bring back RT delalloc support v6 Christoph Hellwig
2024-04-22 11:20 ` [PATCH 11/13] xfs: rework splitting of indirect block reservations Christoph Hellwig

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=20240327151456.GZ6390@frogsfrogsfrogs \
    --to=djwong@kernel.org \
    --cc=chandan.babu@oracle.com \
    --cc=david@fromorbit.com \
    --cc=hch@lst.de \
    --cc=linux-xfs@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