From: "Darrick J. Wong" <djwong@kernel.org>
To: Lukas Herbolt <lukas@herbolt.com>
Cc: linux-xfs@vger.kernel.org, cem@kernel.org, hch@infradead.org,
p.raghav@samsung.com
Subject: Re: [PATCH v11 2/2] xfs: add FALLOC_FL_WRITE_ZEROES to XFS code base
Date: Mon, 9 Mar 2026 17:44:00 -0700 [thread overview]
Message-ID: <20260310004400.GU6033@frogsfrogsfrogs> (raw)
In-Reply-To: <20260309181235.428151-2-lukas@herbolt.com>
On Mon, Mar 09, 2026 at 07:12:36PM +0100, Lukas Herbolt wrote:
> Add support for FALLOC_FL_WRITE_ZEROES if the underlying device
> enable the unmap write zeroes operation.
>
> Co-developed-by: Pankaj Raghav <p.raghav@samsung.com>
> Signed-off-by: Pankaj Raghav <p.raghav@samsung.com>
> Signed-off-by: Lukas Herbolt <lukas@herbolt.com>
>
> ---
> v11 changes:
> - split into 2 patches separating the bmapi_flags addition
> - 2 step allocation, to avoid zeroing beyond EOF
>
> fs/xfs/xfs_file.c | 41 +++++++++++++++++++++++++++++------------
> 1 file changed, 29 insertions(+), 12 deletions(-)
>
> diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
> index fd049a1fc9c6..f8c1611e3267 100644
> --- a/fs/xfs/xfs_file.c
> +++ b/fs/xfs/xfs_file.c
> @@ -1293,29 +1293,45 @@ xfs_falloc_zero_range(
> unsigned int blksize = i_blocksize(inode);
> loff_t new_size = 0;
> int error;
> + bool need_convert = false;
>
> trace_xfs_zero_file_space(ip);
>
> + if (mode & FALLOC_FL_WRITE_ZEROES) {
> + if (xfs_is_always_cow_inode(ip) ||
> + !bdev_write_zeroes_unmap_sectors(
> + xfs_inode_buftarg(ip)->bt_bdev))
> + return -EOPNOTSUPP;
> + need_convert = true;
> + }
> +
> error = xfs_falloc_newsize(file, mode, offset, len, &new_size);
> if (error)
> return error;
>
> if (xfs_falloc_force_zero(ip, ac)) {
> error = xfs_zero_range(ip, offset, len, ac, NULL);
> - } else {
> - error = xfs_free_file_space(ip, offset, len, ac);
> - if (error)
> - return error;
> -
> - len = round_up(offset + len, blksize) -
> - round_down(offset, blksize);
> - offset = round_down(offset, blksize);
> - error = xfs_alloc_file_space(ip, offset, len,
> - XFS_BMAPI_PREALLOC);
> + goto set_filesize;
> }
> + error = xfs_free_file_space(ip, offset, len, ac);
> if (error)
> return error;
> - return xfs_falloc_setsize(file, new_size);
> +
> + len = round_up(offset + len, blksize) - round_down(offset, blksize);
> + offset = round_down(offset, blksize);
> + error = xfs_alloc_file_space(ip, offset, len, XFS_BMAPI_PREALLOC);
> +
> +set_filesize:
> + if (error)
> + return error;
> +
> + error = xfs_falloc_setsize(file, new_size);
> + if (error)
> + return error;
> + if (need_convert)
> + error = xfs_alloc_file_space(ip, offset, len,
> + XFS_BMAPI_CONVERT | XFS_BMAPI_ZERO);
> + return error;
> }
I can't help but think this would be cleaner as:
static int
xfs_falloc_write_zero_range(
struct file *file,
int mode,
loff_t offset,
loff_t len,
struct xfs_zone_alloc_ctx *ac)
{
struct inode *inode = file_inode(file);
struct xfs_inode *ip = XFS_I(inode);
unsigned int blksize = i_blocksize(inode);
loff_t new_size = 0;
int error;
trace_xfs_zero_file_space(ip);
if (xfs_is_always_cow_inode(ip) ||
!bdev_write_zeroes_unmap_sectors(
xfs_inode_buftarg(ip)->bt_bdev))
return -EOPNOTSUPP;
error = xfs_falloc_newsize(file, mode, offset, len, &new_size);
if (error)
return error;
if (xfs_falloc_force_zero(ip, ac)) {
error = xfs_zero_range(ip, offset, len, ac, NULL);
if (error)
return error;
return xfs_falloc_setsize(file, new_size);
}
error = xfs_free_file_space(ip, offset, len, ac);
if (error)
return error;
len = round_up(offset + len, blksize) -
round_down(offset, blksize);
offset = round_down(offset, blksize);
error = xfs_alloc_file_space(ip, offset, len);
if (error)
return error;
error = xfs_falloc_setsize(file, new_size);
if (error)
return error;
return xfs_alloc_file_space(ip, offset, len,
XFS_BMAPI_CONVERT | XFS_BMAPI_ZERO);
}
...assuming there's even a point to WRITE_ZEROES on a zoned file?
--D
>>
> static int
> @@ -1377,7 +1393,7 @@ xfs_falloc_allocate_range(
> (FALLOC_FL_ALLOCATE_RANGE | FALLOC_FL_KEEP_SIZE | \
> FALLOC_FL_PUNCH_HOLE | FALLOC_FL_COLLAPSE_RANGE | \
> FALLOC_FL_ZERO_RANGE | FALLOC_FL_INSERT_RANGE | \
> - FALLOC_FL_UNSHARE_RANGE)
> + FALLOC_FL_UNSHARE_RANGE | FALLOC_FL_WRITE_ZEROES)
>
> STATIC long
> __xfs_file_fallocate(
> @@ -1420,6 +1436,7 @@ __xfs_file_fallocate(
> case FALLOC_FL_INSERT_RANGE:
> error = xfs_falloc_insert_range(file, offset, len);
> break;
> + case FALLOC_FL_WRITE_ZEROES:
> case FALLOC_FL_ZERO_RANGE:
> error = xfs_falloc_zero_range(file, mode, offset, len, ac);
> break;
> --
> 2.53.0
>
>
next prev parent reply other threads:[~2026-03-10 0:44 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-09 18:12 [PATCH v11 2/2] xfs: add FALLOC_FL_WRITE_ZEROES to XFS code base Lukas Herbolt
2026-03-10 0:44 ` Darrick J. Wong [this message]
2026-03-10 10:10 ` Pankaj Raghav (Samsung)
2026-03-10 11:22 ` Lukas Herbolt
2026-03-10 15:02 ` Darrick J. Wong
2026-03-10 10:20 ` Lukas Herbolt
2026-03-10 14:57 ` Darrick J. Wong
2026-03-11 0:12 ` Dave Chinner
2026-03-12 21:36 ` Pankaj Raghav (Samsung)
2026-03-15 23:49 ` Dave Chinner
2026-03-16 7:23 ` Pankaj Raghav
2026-03-16 5:03 ` Lukas Herbolt
2026-03-17 12:20 ` Pankaj Raghav (Samsung)
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=20260310004400.GU6033@frogsfrogsfrogs \
--to=djwong@kernel.org \
--cc=cem@kernel.org \
--cc=hch@infradead.org \
--cc=linux-xfs@vger.kernel.org \
--cc=lukas@herbolt.com \
--cc=p.raghav@samsung.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