All of lore.kernel.org
 help / color / mirror / Atom feed
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);


  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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.