public inbox for linux-xfs@vger.kernel.org
 help / color / mirror / Atom feed
From: Dave Chinner <david@fromorbit.com>
To: Christoph Hellwig <hch@infradead.org>
Cc: xfs@oss.sgi.com
Subject: Re: [PATCH 4/5] xfs: simplify the fallocate path
Date: Mon, 14 Oct 2013 16:04:24 +1100	[thread overview]
Message-ID: <20131014050424.GI4446@dastard> (raw)
In-Reply-To: <20131012075640.467511096@bombadil.infradead.org>

On Sat, Oct 12, 2013 at 12:55:07AM -0700, Christoph Hellwig wrote:
> Call xfs_alloc_file_space or xfs_free_file_space directly from
> xfs_file_fallocate instead of going through xfs_change_file_space.
> 
> This simplified the code by removing the unessecary marshalling of the
> arguments into an xfs_flock64_t structure and allows removing checks that
> are already done in the VFS code.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> 
> ---
>  fs/xfs/xfs_bmap_util.c |   39 ++-----------------------
>  fs/xfs/xfs_bmap_util.h |    4 ++
>  fs/xfs/xfs_file.c      |   76 ++++++++++++++++++++++++++++++-------------------
>  3 files changed, 56 insertions(+), 63 deletions(-)
> 
> Index: xfs/fs/xfs/xfs_file.c
> ===================================================================
> --- xfs.orig/fs/xfs/xfs_file.c	2013-10-01 21:06:23.000000000 +0200
> +++ xfs/fs/xfs/xfs_file.c	2013-10-01 21:12:47.552218689 +0200
> @@ -805,44 +805,64 @@ out:
>  
>  STATIC long
>  xfs_file_fallocate(
.....
> +	tp = xfs_trans_alloc(ip->i_mount, XFS_TRANS_WRITEID);
> +	error = xfs_trans_reserve(tp, &M_RES(ip->i_mount)->tr_writeid, 0, 0);
> +	if (error) {
> +		xfs_trans_cancel(tp, 0);
> +		goto out_unlock;
> +	}
> +
> +	xfs_ilock(ip, XFS_ILOCK_EXCL);
> +	xfs_trans_ijoin(tp, ip, XFS_ILOCK_EXCL);
> +	ip->i_d.di_mode &= ~S_ISUID;
> +	if (ip->i_d.di_mode & S_IXGRP)
> +		ip->i_d.di_mode &= ~S_ISGID;
> +
> +	if (!(mode & FALLOC_FL_PUNCH_HOLE))
> +		ip->i_d.di_flags |= XFS_DIFLAG_PREALLOC;
>  
> -	error = -xfs_change_file_space(ip, cmd, &bf, 0, attr_flags);
> +	xfs_trans_ichgtime(tp, ip, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
> +	xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
> +
> +	if (file->f_flags & O_DSYNC)
> +		xfs_trans_set_sync(tp);
> +	error = xfs_trans_commit(tp, 0);
>  	if (error)
>  		goto out_unlock;

Seems a bit clunky to do all this work when we've got to repeat most
of it when when we call xfs_setattr_size() if the size has changed.
Any thoughts on how we might reduce to a single transaction?

Otherwise it looks ok.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

  reply	other threads:[~2013-10-14  5:05 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-10-12  7:55 [PATCH 0/5] refactor the preallocation and hole punching code Christoph Hellwig
2013-10-12  7:55 ` [PATCH 1/5] xfs: always take the iolock around xfs_setattr_size Christoph Hellwig
2013-10-14  4:50   ` Dave Chinner
2013-10-14  7:29     ` Christoph Hellwig
2013-10-14 14:09   ` [PATCH 1/5 v3] " Christoph Hellwig
2013-10-14 20:28     ` Dave Chinner
2013-10-12  7:55 ` [PATCH 2/5] xfs: remove the unused XFS_ATTR_NONBLOCK flag Christoph Hellwig
2013-10-14  4:51   ` Dave Chinner
2013-10-17 20:01     ` Ben Myers
2013-10-17 20:03       ` Christoph Hellwig
2013-10-17 21:17         ` Ben Myers
2013-10-12  7:55 ` [PATCH 3/5] xfs: always hold the iolock when calling xfs_change_file_space Christoph Hellwig
2013-10-14  4:55   ` Dave Chinner
2013-10-12  7:55 ` [PATCH 4/5] xfs: simplify the fallocate path Christoph Hellwig
2013-10-14  5:04   ` Dave Chinner [this message]
2013-10-14  7:30     ` Christoph Hellwig
2013-10-14 20:03       ` Dave Chinner
2013-10-12  7:55 ` [PATCH 5/5] xfs: fold xfs_change_file_space into xfs_ioc_space Christoph Hellwig
2013-10-14  5:08   ` Dave Chinner
2013-10-15 15:31     ` Christoph Hellwig
2013-10-15 21:47       ` Dave Chinner
2013-10-16  7:03         ` Christoph Hellwig
2013-10-21 21:58 ` [PATCH 0/5] refactor the preallocation and hole punching code Ben Myers
  -- strict thread matches above, loose matches on Subject: below --
2012-12-08 12:08 Christoph Hellwig
2012-12-08 12:08 ` [PATCH 4/5] xfs: simplify the fallocate path Christoph Hellwig
2012-12-10  2:09   ` Dave Chinner
2012-12-10 10:52     ` 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=20131014050424.GI4446@dastard \
    --to=david@fromorbit.com \
    --cc=hch@infradead.org \
    --cc=xfs@oss.sgi.com \
    /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