From: "Darrick J. Wong" <djwong@kernel.org>
To: John Garry <john.g.garry@oracle.com>
Cc: brauner@kernel.org, cem@kernel.org, dchinner@redhat.com,
hch@lst.de, linux-xfs@vger.kernel.org,
linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
ojaswin@linux.ibm.com, ritesh.list@gmail.com,
martin.petersen@oracle.com
Subject: Re: [PATCH RFC 10/10] xfs: Allow block allocator to take an alignment hint
Date: Wed, 5 Feb 2025 11:20:39 -0800 [thread overview]
Message-ID: <20250205192039.GU21808@frogsfrogsfrogs> (raw)
In-Reply-To: <20250204120127.2396727-11-john.g.garry@oracle.com>
On Tue, Feb 04, 2025 at 12:01:27PM +0000, John Garry wrote:
> When issuing an atomic write by the CoW method, give the block allocator a
> hint to naturally align the data blocks.
>
> This means that we have a better chance to issuing the atomic write via
> HW offload next time.
>
> Signed-off-by: John Garry <john.g.garry@oracle.com>
> ---
> fs/xfs/libxfs/xfs_bmap.c | 7 ++++++-
> fs/xfs/libxfs/xfs_bmap.h | 6 +++++-
> fs/xfs/xfs_reflink.c | 8 ++++++--
> 3 files changed, 17 insertions(+), 4 deletions(-)
>
> diff --git a/fs/xfs/libxfs/xfs_bmap.c b/fs/xfs/libxfs/xfs_bmap.c
> index 40ad22fb808b..7a3910018dee 100644
> --- a/fs/xfs/libxfs/xfs_bmap.c
> +++ b/fs/xfs/libxfs/xfs_bmap.c
> @@ -3454,6 +3454,12 @@ xfs_bmap_compute_alignments(
> align = xfs_get_cowextsz_hint(ap->ip);
> else if (ap->datatype & XFS_ALLOC_USERDATA)
> align = xfs_get_extsz_hint(ap->ip);
> +
> + if (align > 1 && ap->flags & XFS_BMAPI_NALIGN)
> + args->alignment = align;
> + else
> + args->alignment = 1;
> +
> if (align) {
> if (xfs_bmap_extsize_align(mp, &ap->got, &ap->prev, align, 0,
> ap->eof, 0, ap->conv, &ap->offset,
> @@ -3781,7 +3787,6 @@ xfs_bmap_btalloc(
> .wasdel = ap->wasdel,
> .resv = XFS_AG_RESV_NONE,
> .datatype = ap->datatype,
> - .alignment = 1,
> .minalignslop = 0,
> };
> xfs_fileoff_t orig_offset;
> diff --git a/fs/xfs/libxfs/xfs_bmap.h b/fs/xfs/libxfs/xfs_bmap.h
> index 4b721d935994..d68b594c3fa2 100644
> --- a/fs/xfs/libxfs/xfs_bmap.h
> +++ b/fs/xfs/libxfs/xfs_bmap.h
> @@ -87,6 +87,9 @@ struct xfs_bmalloca {
> /* Do not update the rmap btree. Used for reconstructing bmbt from rmapbt. */
> #define XFS_BMAPI_NORMAP (1u << 10)
>
> +/* Try to naturally align allocations */
> +#define XFS_BMAPI_NALIGN (1u << 11)
> +
> #define XFS_BMAPI_FLAGS \
> { XFS_BMAPI_ENTIRE, "ENTIRE" }, \
> { XFS_BMAPI_METADATA, "METADATA" }, \
> @@ -98,7 +101,8 @@ struct xfs_bmalloca {
> { XFS_BMAPI_REMAP, "REMAP" }, \
> { XFS_BMAPI_COWFORK, "COWFORK" }, \
> { XFS_BMAPI_NODISCARD, "NODISCARD" }, \
> - { XFS_BMAPI_NORMAP, "NORMAP" }
> + { XFS_BMAPI_NORMAP, "NORMAP" },\
> + { XFS_BMAPI_NALIGN, "NALIGN" }
Tihs isn't really "naturally" aligned, is it? It really means "try to
align allocations to the extent size hint", which isn't required to be a
power of two.
--D
>
>
> static inline int xfs_bmapi_aflag(int w)
> diff --git a/fs/xfs/xfs_reflink.c b/fs/xfs/xfs_reflink.c
> index 60c986300faa..198fb5372f10 100644
> --- a/fs/xfs/xfs_reflink.c
> +++ b/fs/xfs/xfs_reflink.c
> @@ -445,6 +445,11 @@ xfs_reflink_fill_cow_hole(
> int nimaps;
> int error;
> bool found;
> + uint32_t bmapi_flags = XFS_BMAPI_COWFORK |
> + XFS_BMAPI_PREALLOC;
> +
> + if (atomic)
> + bmapi_flags |= XFS_BMAPI_NALIGN;
>
> resaligned = xfs_aligned_fsb_count(imap->br_startoff,
> imap->br_blockcount, xfs_get_cowextsz_hint(ip));
> @@ -478,8 +483,7 @@ xfs_reflink_fill_cow_hole(
> /* Allocate the entire reservation as unwritten blocks. */
> nimaps = 1;
> error = xfs_bmapi_write(tp, ip, imap->br_startoff, imap->br_blockcount,
> - XFS_BMAPI_COWFORK | XFS_BMAPI_PREALLOC, 0, cmap,
> - &nimaps);
> + bmapi_flags, 0, cmap, &nimaps);
> if (error)
> goto out_trans_cancel;
>
> --
> 2.31.1
>
>
next prev parent reply other threads:[~2025-02-05 19:20 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-04 12:01 [PATCH RFC 00/10] large atomic writes for xfs with CoW John Garry
2025-02-04 12:01 ` [PATCH RFC 01/10] xfs: Switch atomic write size check in xfs_file_write_iter() John Garry
2025-02-04 12:01 ` [PATCH RFC 02/10] xfs: Refactor xfs_reflink_end_cow_extent() John Garry
2025-02-05 19:50 ` Darrick J. Wong
2025-02-06 10:35 ` John Garry
2025-02-06 21:38 ` Darrick J. Wong
2025-02-04 12:01 ` [PATCH RFC 03/10] iomap: Support CoW-based atomic writes John Garry
2025-02-05 20:11 ` Darrick J. Wong
2025-02-06 11:21 ` John Garry
2025-02-06 21:40 ` Darrick J. Wong
2025-02-04 12:01 ` [PATCH RFC 04/10] xfs: Make xfs_find_trim_cow_extent() public John Garry
2025-02-04 12:01 ` [PATCH RFC 05/10] xfs: Reflink CoW-based atomic write support John Garry
2025-02-04 12:01 ` [PATCH RFC 06/10] xfs: iomap " John Garry
2025-02-05 20:05 ` Darrick J. Wong
2025-02-06 11:10 ` John Garry
2025-02-06 21:44 ` Darrick J. Wong
2025-02-07 11:48 ` John Garry
2025-02-04 12:01 ` [PATCH RFC 07/10] xfs: Add xfs_file_dio_write_atomic() John Garry
2025-02-05 19:55 ` Darrick J. Wong
2025-02-06 10:43 ` John Garry
2025-02-10 16:59 ` John Garry
2025-02-04 12:01 ` [PATCH RFC 08/10] xfs: Commit CoW-based atomic writes atomically John Garry
2025-02-05 19:47 ` Darrick J. Wong
2025-02-06 10:27 ` John Garry
2025-02-06 21:50 ` Darrick J. Wong
2025-02-07 11:52 ` John Garry
2025-02-04 12:01 ` [PATCH RFC 09/10] xfs: Update atomic write max size John Garry
2025-02-05 19:41 ` Darrick J. Wong
2025-02-06 9:15 ` John Garry
2025-02-06 21:54 ` Darrick J. Wong
2025-02-07 11:53 ` John Garry
2025-02-04 12:01 ` [PATCH RFC 10/10] xfs: Allow block allocator to take an alignment hint John Garry
2025-02-05 19:20 ` Darrick J. Wong [this message]
2025-02-06 8:10 ` John Garry
2025-02-06 21:54 ` Darrick J. Wong
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=20250205192039.GU21808@frogsfrogsfrogs \
--to=djwong@kernel.org \
--cc=brauner@kernel.org \
--cc=cem@kernel.org \
--cc=dchinner@redhat.com \
--cc=hch@lst.de \
--cc=john.g.garry@oracle.com \
--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 \
/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