From: John Garry <john.g.garry@oracle.com>
To: Dave Chinner <david@fromorbit.com>
Cc: djwong@kernel.org, hch@lst.de, viro@zeniv.linux.org.uk,
brauner@kernel.org, jack@suse.cz, chandan.babu@oracle.com,
axboe@kernel.dk, martin.petersen@oracle.com,
linux-kernel@vger.kernel.org, linux-xfs@vger.kernel.org,
linux-fsdevel@vger.kernel.org, tytso@mit.edu, jbongio@google.com,
ojaswin@linux.ibm.com, ritesh.list@gmail.com,
linux-block@vger.kernel.org
Subject: Re: [PATCH v2 06/14] fs: xfs: Do not free EOF blocks for forcealign
Date: Thu, 7 Mar 2024 11:38:54 +0000 [thread overview]
Message-ID: <dac177e5-793a-4955-bc82-9838a1ce6dd5@oracle.com> (raw)
In-Reply-To: <ZejbFuavNva4ut/3@dread.disaster.area>
>> */
>> end_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)XFS_ISIZE(ip));
>> - if (XFS_IS_REALTIME_INODE(ip) && mp->m_sb.sb_rextsize > 1)
>> +
>> + /* Do not free blocks when forcing extent sizes */
>
> That comment seems wrong - this just rounds up where we start
> trimming from to be aligned...
ok
>
>> + if (xfs_get_extsz(ip) > 1)
>> + end_fsb = roundup_64(end_fsb, xfs_get_extsz(ip));
>> + else if (XFS_IS_REALTIME_INODE(ip) && mp->m_sb.sb_rextsize > 1)
>> end_fsb = xfs_rtb_roundup_rtx(mp, end_fsb);
>
> I think this would be better written as:
>
> /*
> * Forced extent alignment requires us to round up where we
> * start trimming from so that result is correctly aligned.
> */
> if (xfs_inode_forcealign(ip)) {
> if (ip->i_extsize > 1)
> end_fsb = roundup_64(end_fsb, ip->i_extsize);
> else if (XFS_IS_REALTIME_INODE(ip) && mp->m_sb.sb_rextsize > 1)
> end_fsb = xfs_rtb_roundup_rtx(mp, end_fsb);
> }
>
> Because we only want end-fsb alignment when forced alignment is
> required.
But why change current rtvol behaviour?
>
> Which then makes me wonder: truncate needs this, too, doesn't it?
> And the various fallocate() ops like hole punching and extent
> shifting?
>
Yes, I would think so. I quickly checked rtvol for truncate and it does
the round up. I would need to check the relevant code for truncate and
fallocate for forcealign now.
I do also wonder if we could factor out this rounding up code for
truncate, facallocate, and whatever else.
>> diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c
>> index 2c439df8c47f..bbb8886f1d32 100644
>> --- a/fs/xfs/xfs_inode.c
>> +++ b/fs/xfs/xfs_inode.c
>> @@ -65,6 +65,20 @@ xfs_get_extsz_hint(
>> return 0;
>> }
>>
>> +/*
>> + * Helper function to extract extent size. It will return a power-of-2,
>> + * as forcealign requires this.
>> + */
>> +xfs_extlen_t
>> +xfs_get_extsz(
>> + struct xfs_inode *ip)
>> +{
>> + if (xfs_inode_forcealign(ip) && ip->i_extsize)
>> + return ip->i_extsize;
>> +
>> + return 1;
>> +}
>
> This can go away - if it is needed elsewhere, then I think it would
> be better open coded because it better documents what the code is
> doing...
>
I would rather get rid of xfs_get_extsz() for sure.
Thanks,
John
next prev parent reply other threads:[~2024-03-07 11:40 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-03-04 13:04 [PATCH v2 00/14] block atomic writes for XFS John Garry
2024-03-04 13:04 ` [PATCH v2 01/14] block: Add blk_validate_atomic_write_op_size() John Garry
2024-03-04 13:04 ` [PATCH v2 02/14] fs: xfs: Don't use low-space allocator for alignment > 1 John Garry
2024-03-04 22:15 ` Dave Chinner
2024-03-05 13:36 ` John Garry
2024-03-04 13:04 ` [PATCH v2 03/14] fs: xfs: Introduce FORCEALIGN inode flag John Garry
2024-03-04 13:04 ` [PATCH v2 04/14] fs: xfs: Make file data allocations observe the 'forcealign' flag John Garry
2024-03-05 0:44 ` Dave Chinner
2024-03-05 15:22 ` John Garry
2024-03-05 22:18 ` Dave Chinner
2024-03-06 9:41 ` John Garry
2024-03-04 13:04 ` [PATCH v2 05/14] fs: xfs: Enable file data forcealign feature John Garry
2024-03-04 13:04 ` [PATCH v2 06/14] fs: xfs: Do not free EOF blocks for forcealign John Garry
2024-03-06 21:07 ` Dave Chinner
2024-03-07 11:38 ` John Garry [this message]
2024-03-04 13:04 ` [PATCH v2 07/14] fs: iomap: Sub-extent zeroing John Garry
2024-03-06 21:14 ` Dave Chinner
2024-03-07 11:51 ` John Garry
2024-03-04 13:04 ` [PATCH v2 08/14] fs: xfs: " John Garry
2024-03-06 22:00 ` Dave Chinner
2024-03-07 12:57 ` John Garry
2024-03-04 13:04 ` [PATCH v2 09/14] fs: Add FS_XFLAG_ATOMICWRITES flag John Garry
2024-03-04 13:04 ` [PATCH v2 10/14] fs: iomap: Atomic write support John Garry
2024-03-04 13:04 ` [PATCH v2 11/14] fs: xfs: Support FS_XFLAG_ATOMICWRITES for forcealign John Garry
2024-03-06 21:43 ` Dave Chinner
2024-03-07 12:42 ` John Garry
2024-03-04 13:04 ` [PATCH v2 12/14] fs: xfs: Support atomic write for statx John Garry
2024-03-06 21:31 ` Dave Chinner
2024-03-07 10:35 ` John Garry
2024-03-04 13:04 ` [PATCH v2 13/14] fs: xfs: Validate atomic writes John Garry
2024-03-06 21:22 ` Dave Chinner
2024-03-07 10:19 ` John Garry
2024-03-04 13:04 ` [PATCH v2 14/14] fs: xfs: Support setting FMODE_CAN_ATOMIC_WRITE John Garry
2024-03-06 21:33 ` Dave Chinner
2024-03-07 11:55 ` John Garry
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=dac177e5-793a-4955-bc82-9838a1ce6dd5@oracle.com \
--to=john.g.garry@oracle.com \
--cc=axboe@kernel.dk \
--cc=brauner@kernel.org \
--cc=chandan.babu@oracle.com \
--cc=david@fromorbit.com \
--cc=djwong@kernel.org \
--cc=hch@lst.de \
--cc=jack@suse.cz \
--cc=jbongio@google.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
--cc=martin.petersen@oracle.com \
--cc=ojaswin@linux.ibm.com \
--cc=ritesh.list@gmail.com \
--cc=tytso@mit.edu \
--cc=viro@zeniv.linux.org.uk \
/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).