From: "Darrick J. Wong" <djwong@kernel.org>
To: Joanne Koong <joannelkoong@gmail.com>
Cc: brauner@kernel.org, hch@lst.de, willy@infradead.org,
hsiangkao@linux.alibaba.com, linux-fsdevel@vger.kernel.org,
linux-xfs@vger.kernel.org,
open list <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v2 01/18] iomap: add ->iomap_next() and iomap_process() helper
Date: Thu, 2 Jul 2026 09:23:08 -0700 [thread overview]
Message-ID: <20260702162308.GH9392@frogsfrogsfrogs> (raw)
In-Reply-To: <20260701000949.1666714-2-joannelkoong@gmail.com>
On Tue, Jun 30, 2026 at 05:09:16PM -0700, Joanne Koong wrote:
> Have one ->iomap_next() callback instead of ->iomap_begin() and
> ->iomap_end(). ->iomap_next() finishes the previous mapping if needed,
> and produces the next mapping. This lets performance-critical callers
> inline the iteration with a fixed callback, which the compiler is able
> to call directly instead of indirectly.
Er... what is being called directly here? The iomap_{begin,end}
functions? It looks to me like ->iomap_next is still an indirect call
even at the end of the series, right? That's still a net reduction of
indirect calls at least.
Oh, wait, it's the __always_inline iomap_process function that a smart
compiler can use to turn the indirect calls into direct ones, isn't it?
It might be useful to add a comment to that function saying that out
loud so that dolts like me will pick up on why it's critical for it to
be an inline function.
I like how this is going. :)
> iomap_iter() uses ->iomap_next() when the filesystem provides that
> callback and otherwise falls back to the ->iomap_begin()/->iomap_end()
> path, so filesystems can be converted one at a time.
> Add a iomap_process() inline helper that does most of the logic needed
> in an ->iomap_next() implementation.
>
> Suggested-by: Christoph Hellwig <hch@lst.de>
> Suggested-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
> ---
> fs/iomap/iter.c | 113 ++++++++++++++++++++++++++++++++++++------
> include/linux/iomap.h | 91 +++++++++++++++++++++++++++-------
> 2 files changed, 171 insertions(+), 33 deletions(-)
>
> diff --git a/fs/iomap/iter.c b/fs/iomap/iter.c
> index e4a29829591a..1062e4e34c38 100644
> --- a/fs/iomap/iter.c
> +++ b/fs/iomap/iter.c
> @@ -39,22 +39,7 @@ static inline void iomap_iter_done(struct iomap_iter *iter)
> trace_iomap_iter_srcmap(iter->inode, &iter->srcmap);
> }
>
> -/**
> - * iomap_iter - iterate over a ranges in a file
> - * @iter: iteration structue
> - * @ops: iomap ops provided by the file system
> - *
> - * Iterate over filesystem-provided space mappings for the provided file range.
> - *
> - * This function handles cleanup of resources acquired for iteration when the
> - * filesystem indicates there are no more space mappings, which means that this
> - * function must be called in a loop that continues as long it returns a
> - * positive value. If 0 or a negative value is returned, the caller must not
> - * return to the loop body. Within a loop body, there are two ways to break out
> - * of the loop body: leave @iter.status unchanged, or set it to a negative
> - * errno.
> - */
> -int iomap_iter(struct iomap_iter *iter, const struct iomap_ops *ops)
> +static int iomap_iter_legacy(struct iomap_iter *iter, const struct iomap_ops *ops)
> {
> bool stale = iter->iomap.flags & IOMAP_F_STALE;
> ssize_t advanced;
> @@ -114,3 +99,99 @@ int iomap_iter(struct iomap_iter *iter, const struct iomap_ops *ops)
> iomap_iter_done(iter);
> return 1;
> }
> +
> +static int iomap_iter_next(struct iomap_iter *iter, const struct iomap_ops *ops)
> +{
> + int ret;
> +
> + trace_iomap_iter(iter, ops, _RET_IP_);
> +
> + ret = ops->iomap_next(iter, &iter->iomap, &iter->srcmap);
> + iter->status = 0;
> + if (ret > 0)
> + iomap_iter_done(iter);
> +
> + return ret;
> +}
> +
> +/**
> + * iomap_iter - iterate over a ranges in a file
> + * @iter: iteration structue
> + * @ops: iomap ops provided by the file system
> + *
> + * Iterate over filesystem-provided space mappings for the provided file range.
> + *
> + * This function handles cleanup of resources acquired for iteration when the
> + * filesystem indicates there are no more space mappings, which means that this
> + * function must be called in a loop that continues as long it returns a
> + * positive value. If 0 or a negative value is returned, the caller must not
> + * return to the loop body. Within a loop body, there are two ways to break out
> + * of the loop body: leave @iter.status unchanged, or set it to a negative
> + * errno.
> + */
> +int iomap_iter(struct iomap_iter *iter, const struct iomap_ops *ops)
> +{
> + if (ops->iomap_next)
> + return iomap_iter_next(iter, ops);
> +
> + return iomap_iter_legacy(iter, ops);
> +}
> +
> +/**
> + * iomap_iter_continue - decide whether iteration should continue
> + * @iter: iteration structure
> + * @iomap: the mapping that was just processed
> + * @srcmap: the source mapping that was just processed
> + *
> + * Helper for ->iomap_next() implementations, normally called via
> + * iomap_process(). Called after the previous mapping has been finished to
> + * determine whether there is more of the file range left to process.
> + *
> + * Returns 1 if there is more work to do, in which case @iomap and @srcmap are
> + * cleared so the caller can produce the next mapping; zero if the range is
> + * fully consumed; or a negative errno on error. Any folio batch attached to
> + * the mapping is released before returning.
> + */
> +int iomap_iter_continue(const struct iomap_iter *iter, struct iomap *iomap,
> + struct iomap *srcmap, int ret)
> +{
> + bool stale = iomap->flags & IOMAP_F_STALE;
> + ssize_t advanced = iter->pos - iter->iter_start_pos;
These both could be const, right?
--D
> +
> + if (!iomap->length)
> + return 1;
> +
> + /*
> + * Use iter->len to determine whether to continue onto the next mapping.
> + * Explicitly terminate on error status or if the current iter has not
> + * advanced at all (i.e. no work was done for some reason) unless the
> + * mapping has been marked stale and needs to be reprocessed.
> + */
> + if (ret < 0 && !advanced)
> + return ret;
> +
> + /* detect old return semantics where this would advance */
> + if (WARN_ON_ONCE(iter->status > 0))
> + ret = -EIO;
> + else if (iter->status < 0)
> + ret = iter->status;
> + else if (iter->len == 0 || (!advanced && !stale))
> + ret = 0;
> + else
> + ret = 1;
> +
> + if (iomap->flags & IOMAP_F_FOLIO_BATCH) {
> + folio_batch_release(iter->fbatch);
> + folio_batch_reinit(iter->fbatch);
> + iomap->flags &= ~IOMAP_F_FOLIO_BATCH;
> + }
> +
> + if (ret <= 0)
> + return ret;
> +
> + memset(iomap, 0, sizeof(*iomap));
> + memset(srcmap, 0, sizeof(*srcmap));
> +
> + return ret;
> +}
> +EXPORT_SYMBOL_GPL(iomap_iter_continue);
> diff --git a/include/linux/iomap.h b/include/linux/iomap.h
> index 3582ed1fe236..8a78f47c557b 100644
> --- a/include/linux/iomap.h
> +++ b/include/linux/iomap.h
> @@ -212,24 +212,35 @@ struct iomap_write_ops {
> #define IOMAP_ATOMIC (1 << 9) /* torn-write protection */
> #define IOMAP_DONTCACHE (1 << 10)
>
> -struct iomap_ops {
> - /*
> - * Return the existing mapping at pos, or reserve space starting at
> - * pos for up to length, as long as we can do it as a single mapping.
> - * The actual length is returned in iomap->length.
> - */
> - int (*iomap_begin)(struct inode *inode, loff_t pos, loff_t length,
> - unsigned flags, struct iomap *iomap,
> - struct iomap *srcmap);
> +/*
> + * Return the existing mapping at pos, or reserve space starting at pos for up
> + * to length, as long as we can do it as a single mapping.
> + * The actual length is returned in iomap->length.
> + */
> +typedef int (*iomap_begin_fn)(struct inode *inode, loff_t pos, loff_t length,
> + unsigned flags, struct iomap *iomap, struct iomap *srcmap);
>
> - /*
> - * Commit and/or unreserve space previous allocated using iomap_begin.
> - * Written indicates the length of the successful write operation which
> - * needs to be commited, while the rest needs to be unreserved.
> - * Written might be zero if no data was written.
> - */
> - int (*iomap_end)(struct inode *inode, loff_t pos, loff_t length,
> - ssize_t written, unsigned flags, struct iomap *iomap);
> +/*
> + * Commit and/or unreserve space previous allocated using iomap_begin.
> + * Written indicates the length of the successful write operation which needs
> + * to be commited, while the rest needs to be unreserved.
> + * Written might be zero if no data was written.
> + */
> +typedef int (*iomap_end_fn)(struct inode *inode, loff_t pos, loff_t length,
> + ssize_t written, unsigned flags, struct iomap *iomap);
> +
> +/*
> + * Produce the next mapping (finishing the previous one if needed).
> + * Return 1 to continue iterating, 0 if the range is fully consumed, or a
> + * negative error on failure.
> + */
> +typedef int (*iomap_next_fn)(const struct iomap_iter *iter, struct iomap *iomap,
> + struct iomap *srcmap);
> +
> +struct iomap_ops {
> + iomap_begin_fn iomap_begin;
> + iomap_end_fn iomap_end;
> + iomap_next_fn iomap_next;
> };
>
> /**
> @@ -317,6 +328,9 @@ static inline const struct iomap *iomap_iter_srcmap(const struct iomap_iter *i)
> return &i->iomap;
> }
>
> +int iomap_iter_continue(const struct iomap_iter *iter, struct iomap *iomap,
> + struct iomap *srcmap, int ret);
> +
> /*
> * Return the file offset for the first unchanged block after a short write.
> *
> @@ -648,4 +662,47 @@ static inline void iomap_bio_readahead(struct readahead_control *rac,
> }
> #endif /* CONFIG_BLOCK */
>
> +/**
> + * iomap_process - finish the previous mapping and produce the next one
> + * @iter: iteration structure
> + * @iomap: mapping to finish and then repopulate
> + * @srcmap: source mapping to finish and then repopulate
> + * @begin: callback that produces a mapping for the current position
> + * @end: optional callback that finishes the previous mapping, or NULL
> + *
> + * Inline helper that implements the common body of an ->iomap_next()
> + * callback: it finishes the previous mapping via @end (if present), decides
> + * via iomap_iter_continue() whether to keep going, and obtains the next
> + * mapping via @begin.
> + *
> + * Returns 1 to continue iterating, 0 once the range is fully consumed, or a
> + * negative errno on error.
> + */
> +static __always_inline int iomap_process(const struct iomap_iter *iter,
> + struct iomap *iomap, struct iomap *srcmap, iomap_begin_fn begin,
> + iomap_end_fn end)
> +{
> + int ret = 0;
> +
> + if (iomap->length && end) {
> + ssize_t advanced = iter->pos - iter->iter_start_pos;
> + loff_t len;
> +
> + len = iomap_length_trim(iter, iter->iter_start_pos,
> + iter->len + advanced);
> +
> + ret = end(iter->inode, iter->iter_start_pos, len, advanced,
> + iter->flags, iomap);
> + }
> +
> + ret = iomap_iter_continue(iter, iomap, srcmap, ret);
> + if (ret <= 0)
> + return ret;
> +
> + ret = begin(iter->inode, iter->pos, iter->len, iter->flags, iomap,
> + srcmap);
> +
> + return ret < 0 ? ret : 1;
> +}
> +
> #endif /* LINUX_IOMAP_H */
> --
> 2.52.0
>
>
next prev parent reply other threads:[~2026-07-02 16:23 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-01 0:09 [PATCH v2 00/18] iomap: convert to in-iter iomap_next() model Joanne Koong
2026-07-01 0:09 ` [PATCH v2 01/18] iomap: add ->iomap_next() and iomap_process() helper Joanne Koong
2026-07-02 14:00 ` Christoph Hellwig
2026-07-02 23:01 ` Joanne Koong
2026-07-02 16:23 ` Darrick J. Wong [this message]
2026-07-02 22:41 ` Joanne Koong
2026-07-01 0:09 ` [PATCH v2 02/18] xfs: convert iomap ops to ->iomap_next() Joanne Koong
2026-07-02 14:03 ` Christoph Hellwig
2026-07-02 16:48 ` Darrick J. Wong
2026-07-02 16:43 ` Darrick J. Wong
2026-07-02 23:59 ` Joanne Koong
2026-07-03 1:21 ` Joanne Koong
2026-07-03 1:41 ` Darrick J. Wong
2026-07-03 1:38 ` Darrick J. Wong
2026-07-01 0:09 ` [PATCH v2 03/18] btrfs: " Joanne Koong
2026-07-01 0:09 ` [PATCH v2 04/18] ntfs3: " Joanne Koong
2026-07-01 0:09 ` [PATCH v2 05/18] ntfs: " Joanne Koong
2026-07-01 0:09 ` [PATCH v2 06/18] ext4: " Joanne Koong
2026-07-01 10:01 ` Jan Kara
2026-07-01 0:09 ` [PATCH v2 07/18] erofs: " Joanne Koong
2026-07-01 0:41 ` Gao Xiang
2026-07-01 0:09 ` [PATCH v2 08/18] zonefs: " Joanne Koong
2026-07-01 0:09 ` [PATCH v2 09/18] ext2: " Joanne Koong
2026-07-01 10:02 ` Jan Kara
2026-07-01 0:09 ` [PATCH v2 10/18] block: " Joanne Koong
2026-07-01 0:55 ` Keith Busch
2026-07-02 14:03 ` Christoph Hellwig
2026-07-03 0:06 ` Joanne Koong
2026-07-01 0:09 ` [PATCH v2 11/18] f2fs: " Joanne Koong
2026-07-01 0:09 ` [PATCH v2 12/18] gfs2: " Joanne Koong
2026-07-01 0:09 ` [PATCH v2 13/18] hpfs: " Joanne Koong
2026-07-01 0:09 ` [PATCH v2 14/18] fuse: " Joanne Koong
2026-07-01 0:09 ` [PATCH v2 15/18] exfat: " Joanne Koong
2026-07-03 1:41 ` Joanne Koong
2026-07-01 0:09 ` [PATCH v2 16/18] iomap: remove ->iomap_begin()/->iomap_end() legacy path Joanne Koong
2026-07-02 14:04 ` Christoph Hellwig
2026-07-02 16:51 ` Darrick J. Wong
2026-07-01 0:09 ` [PATCH v2 17/18] iomap: pass iomap_next_fn directly instead of struct iomap_ops Joanne Koong
2026-07-01 10:04 ` Jan Kara
2026-07-02 14:07 ` Christoph Hellwig
2026-07-02 16:51 ` Darrick J. Wong
2026-07-03 1:47 ` Joanne Koong
2026-07-03 2:01 ` Darrick J. Wong
2026-07-02 16:58 ` Darrick J. Wong
2026-07-03 0:17 ` Joanne Koong
2026-07-03 1:42 ` Darrick J. Wong
2026-07-01 0:09 ` [PATCH v2 18/18] Documentation: iomap: update docs to reflect iomap_next model Joanne Koong
2026-07-02 14:07 ` Christoph Hellwig
2026-07-02 19:26 ` Darrick J. Wong
2026-07-03 1:36 ` Joanne Koong
2026-07-03 2:00 ` Darrick J. Wong
2026-07-02 13:49 ` [PATCH v2 00/18] iomap: convert to in-iter iomap_next() model Christoph Hellwig
2026-07-03 1:08 ` Joanne Koong
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=20260702162308.GH9392@frogsfrogsfrogs \
--to=djwong@kernel.org \
--cc=brauner@kernel.org \
--cc=hch@lst.de \
--cc=hsiangkao@linux.alibaba.com \
--cc=joannelkoong@gmail.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
--cc=willy@infradead.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