From: David Howells <dhowells@redhat.com>
To: Matthew Wilcox <willy@infradead.org>,
Christoph Hellwig <hch@infradead.org>,
Jens Axboe <axboe@kernel.dk>, Leon Romanovsky <leon@kernel.org>
Cc: David Howells <dhowells@redhat.com>,
Christian Brauner <christian@brauner.io>,
Paulo Alcantara <pc@manguebit.com>,
netfs@lists.linux.dev, linux-afs@lists.infradead.org,
linux-cifs@vger.kernel.org, linux-nfs@vger.kernel.org,
ceph-devel@vger.kernel.org, v9fs@lists.linux.dev,
linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
Paulo Alcantara <pc@manguebit.org>,
Steve French <sfrench@samba.org>
Subject: [RFC PATCH 07/17] netfs: Add a function to extract from an iter into a bvecq
Date: Wed, 4 Mar 2026 14:03:14 +0000 [thread overview]
Message-ID: <20260304140328.112636-8-dhowells@redhat.com> (raw)
In-Reply-To: <20260304140328.112636-1-dhowells@redhat.com>
Add a function to extract a slice of data from an iterator of any type into
a bvec queue chain.
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Paulo Alcantara <pc@manguebit.org>
cc: Matthew Wilcox <willy@infradead.org>
cc: Christoph Hellwig <hch@infradead.org>
cc: Steve French <sfrench@samba.org>
cc: linux-cifs@vger.kernel.org
cc: netfs@lists.linux.dev
cc: linux-fsdevel@vger.kernel.org
---
fs/netfs/iterator.c | 122 ++++++++++++++++++++++++++++++++++++++++++
include/linux/netfs.h | 3 ++
2 files changed, 125 insertions(+)
diff --git a/fs/netfs/iterator.c b/fs/netfs/iterator.c
index 72a435e5fc6d..faf4f0a3b33d 100644
--- a/fs/netfs/iterator.c
+++ b/fs/netfs/iterator.c
@@ -13,6 +13,128 @@
#include <linux/netfs.h>
#include "internal.h"
+/**
+ * netfs_extract_iter - Extract the pages from an iterator into a bvecq
+ * @orig: The original iterator
+ * @orig_len: The amount of iterator to copy
+ * @max_segs: Maximum number of contiguous segments
+ * @fpos: Starting file position to label the bvecq with
+ * @_bvecq_head: Where to cache the bvec queue
+ * @extraction_flags: Flags to qualify the request
+ *
+ * Extract the page fragments from the given amount of the source iterator and
+ * build bvec queue that refers to all of those bits. This allows the original
+ * iterator to disposed of.
+ *
+ * @extraction_flags can have ITER_ALLOW_P2PDMA set to request peer-to-peer DMA be
+ * allowed on the pages extracted.
+ *
+ * On success, the amount of data in the bvec is returned, the original
+ * iterator will have been advanced by the amount extracted.
+ *
+ * The bvecq segments are marked with indications on how to get clean up the
+ * extracted fragments.
+ */
+ssize_t netfs_extract_iter(struct iov_iter *orig, size_t orig_len, size_t max_segs,
+ unsigned long long fpos, struct bvecq **_bvecq_head,
+ iov_iter_extraction_t extraction_flags)
+{
+ struct bvecq *bq_tail = NULL;
+ ssize_t ret = 0;
+ size_t segs_per_bq;
+ size_t extracted = 0;
+
+ _enter("{%u,%zx},%zx", orig->iter_type, orig->count, orig_len);
+
+ if (max_segs == 0)
+ max_segs = ULONG_MAX;
+
+ /* We want the biggest pow-of-2 size that has at most 255 segs and that
+ * won't exceed a 4K page.
+ */
+ segs_per_bq = (4096 - sizeof(*bq_tail)) / sizeof(bq_tail->__bv[0]);
+ if (segs_per_bq > 255)
+ segs_per_bq = (2048 - sizeof(*bq_tail)) / sizeof(bq_tail->__bv[0]);
+
+ do {
+ struct bvecq *bq;
+ size_t nr_slots = iov_iter_npages(orig, umin(segs_per_bq, max_segs));
+
+ if (WARN_ON(nr_slots == 0 && extracted < orig_len) ||
+ WARN_ON(nr_slots > max_segs))
+ break;
+ max_segs -= nr_slots;
+
+ bq = netfs_alloc_one_bvecq(nr_slots, GFP_NOFS);
+ if (!bq) {
+ ret = -ENOMEM;
+ break;
+ }
+ bq->free = user_backed_iter(orig);
+ bq->unpin = iov_iter_extract_will_pin(orig);
+ bq->prev = bq_tail;
+ bq->fpos = fpos + extracted;
+
+ if (bq_tail)
+ bq_tail->next = bq;
+ else
+ *_bvecq_head = bq;
+ bq_tail = bq;
+
+ if (extracted >= orig_len)
+ break;
+
+ /* Put the page list at the end of the bvec list storage. bvec
+ * elements are larger than page pointers, so as long as we
+ * work 0->last, we should be fine.
+ */
+ struct bio_vec *bv = bq->bv;
+ struct page **pages;
+ size_t bv_size = array_size(bq->max_segs, sizeof(*bv));
+ size_t pg_size = array_size(bq->max_segs, sizeof(*pages));
+
+ pages = (void *)bv + bv_size - pg_size;
+
+ do {
+ unsigned int cur_npages;
+ ssize_t got;
+ size_t offset;
+
+ got = iov_iter_extract_pages(orig, &pages, orig_len - extracted,
+ bq->max_segs - bq->nr_segs,
+ extraction_flags, &offset);
+ if (got < 0) {
+ pr_err("Couldn't get user pages (rc=%zd)\n", got);
+ ret = got;
+ break;
+ }
+
+ if (got > orig_len - extracted) {
+ pr_err("get_pages rc=%zd more than %zu\n",
+ got, orig_len - extracted);
+ break;
+ }
+
+ extracted += got;
+ got += offset;
+ cur_npages = DIV_ROUND_UP(got, PAGE_SIZE);
+
+ for (unsigned int i = 0; i < cur_npages; i++) {
+ size_t len = umin(got, PAGE_SIZE);
+
+ bvec_set_page(&bq->bv[bq->nr_segs],
+ *pages++, len - offset, offset);
+ bq->nr_segs++;
+ got -= len;
+ offset = 0;
+ }
+ } while (extracted < orig_len && !bvecq_is_full(bq));
+ } while (extracted < orig_len && max_segs > 0);
+
+ return extracted ?: ret;
+}
+EXPORT_SYMBOL_GPL(netfs_extract_iter);
+
/**
* netfs_extract_user_iter - Extract the pages from a user iterator into a bvec
* @orig: The original iterator
diff --git a/include/linux/netfs.h b/include/linux/netfs.h
index f9ad067a0a0c..b146aeaaf6c9 100644
--- a/include/linux/netfs.h
+++ b/include/linux/netfs.h
@@ -448,6 +448,9 @@ void netfs_get_subrequest(struct netfs_io_subrequest *subreq,
enum netfs_sreq_ref_trace what);
void netfs_put_subrequest(struct netfs_io_subrequest *subreq,
enum netfs_sreq_ref_trace what);
+ssize_t netfs_extract_iter(struct iov_iter *orig, size_t orig_len, size_t max_segs,
+ unsigned long long fpos, struct bvecq **_bvecq_head,
+ iov_iter_extraction_t extraction_flags);
ssize_t netfs_extract_user_iter(struct iov_iter *orig, size_t orig_len,
struct iov_iter *new,
iov_iter_extraction_t extraction_flags);
next prev parent reply other threads:[~2026-03-04 14:04 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-04 14:03 [RFC PATCH 00/17] netfs: [WIP] Keep track of folios in a segmented bio_vec[] chain David Howells
2026-03-04 14:03 ` [RFC PATCH 01/17] netfs: Fix unbuffered/DIO writes to dispatch subrequests in strict sequence David Howells
2026-03-04 14:03 ` [RFC PATCH 02/17] vfs: Implement a FIEMAP callback David Howells
2026-03-04 14:06 ` Christoph Hellwig
2026-03-04 14:21 ` David Howells
2026-03-04 14:25 ` Christoph Hellwig
2026-03-04 14:34 ` David Howells
2026-03-04 14:03 ` [RFC PATCH 03/17] iov_iter: Add a segmented queue of bio_vec[] David Howells
2026-03-04 14:03 ` [RFC PATCH 04/17] Add a function to kmap one page of a multipage bio_vec David Howells
2026-03-04 14:03 ` [RFC PATCH 05/17] netfs: Add some tools for managing bvecq chains David Howells
2026-03-04 14:03 ` [RFC PATCH 06/17] afs: Use a bvecq to hold dir content rather than folioq David Howells
2026-03-04 14:03 ` David Howells [this message]
2026-03-04 14:03 ` [RFC PATCH 08/17] cifs: Use a bvecq for buffering instead of a folioq David Howells
2026-03-04 14:03 ` [RFC PATCH 09/17] cifs: Support ITER_BVECQ in smb_extract_iter_to_rdma() David Howells
2026-03-04 14:03 ` [RFC PATCH 10/17] netfs: Switch to using bvecq rather than folio_queue and rolling_buffer David Howells
2026-03-04 14:03 ` [RFC PATCH 11/17] cifs: Remove support for ITER_KVEC/BVEC/FOLIOQ from smb_extract_iter_to_rdma() David Howells
2026-03-04 14:03 ` [RFC PATCH 12/17] netfs: Remove netfs_alloc/free_folioq_buffer() David Howells
2026-03-04 14:03 ` [RFC PATCH 13/17] netfs: Remove netfs_extract_user_iter() David Howells
2026-03-04 14:03 ` [RFC PATCH 14/17] iov_iter: Remove ITER_FOLIOQ David Howells
2026-03-04 14:03 ` [RFC PATCH 15/17] netfs: Remove folio_queue and rolling_buffer David Howells
2026-03-04 14:03 ` [RFC PATCH 16/17] netfs: Check for too much data being read David Howells
2026-03-04 14:03 ` [RFC PATCH 17/17] netfs: Combine prepare and issue ops and grab the buffers on request David Howells
2026-03-04 14:39 ` Christoph Hellwig
2026-03-04 14:51 ` David Howells
2026-03-04 15:01 ` Christoph Hellwig
2026-03-23 18:37 ` ChenXiaoSong
2026-03-23 20:14 ` David Howells
2026-03-23 22:44 ` Paulo Alcantara
2026-03-24 1:03 ` ChenXiaoSong
2026-03-24 7:16 ` David Howells
2026-03-24 7:38 ` ChenXiaoSong
2026-03-24 7:53 ` David Howells
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=20260304140328.112636-8-dhowells@redhat.com \
--to=dhowells@redhat.com \
--cc=axboe@kernel.dk \
--cc=ceph-devel@vger.kernel.org \
--cc=christian@brauner.io \
--cc=hch@infradead.org \
--cc=leon@kernel.org \
--cc=linux-afs@lists.infradead.org \
--cc=linux-cifs@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-nfs@vger.kernel.org \
--cc=netfs@lists.linux.dev \
--cc=pc@manguebit.com \
--cc=pc@manguebit.org \
--cc=sfrench@samba.org \
--cc=v9fs@lists.linux.dev \
--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