public inbox for linux-nvme@lists.infradead.org
 help / color / mirror / Atom feed
From: Keith Busch <kbusch@fb.com>
To: <linux-fsdevel@vger.kernel.org>, <linux-block@vger.kernel.org>,
	<linux-nvme@lists.infradead.org>
Cc: <axboe@kernel.dk>, Kernel Team <Kernel-team@fb.com>, <hch@lst.de>,
	<bvanassche@acm.org>, <damien.lemoal@opensource.wdc.com>,
	<ebiggers@kernel.org>, <pankydev8@gmail.com>,
	Keith Busch <kbusch@kernel.org>
Subject: [PATCHv6 10/11] block: relax direct io memory alignment
Date: Fri, 10 Jun 2022 12:58:29 -0700	[thread overview]
Message-ID: <20220610195830.3574005-11-kbusch@fb.com> (raw)
In-Reply-To: <20220610195830.3574005-1-kbusch@fb.com>

From: Keith Busch <kbusch@kernel.org>

Use the address alignment requirements from the block_device for direct
io instead of requiring addresses be aligned to the block size. User
space can discover the alignment requirements from the dma_alignment
queue attribute.

User space can specify any hardware compatible DMA offset for each
segment, but every segment length is still required to be a multiple of
the block size.

Signed-off-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 block/bio.c            | 9 +++++++++
 block/fops.c           | 4 ++--
 include/linux/blkdev.h | 5 +++++
 3 files changed, 16 insertions(+), 2 deletions(-)

diff --git a/block/bio.c b/block/bio.c
index 5618c6a4b3a3..551f1d12208b 100644
--- a/block/bio.c
+++ b/block/bio.c
@@ -1220,7 +1220,16 @@ static int __bio_iov_iter_get_pages(struct bio *bio, struct iov_iter *iter)
 	BUILD_BUG_ON(PAGE_PTRS_PER_BVEC < 2);
 	pages += entries_left * (PAGE_PTRS_PER_BVEC - 1);
 
+	/*
+	 * Each segment in the iov is required to be a block size multiple.
+	 * However, we may not be able to get the entire segment if it spans
+	 * more pages than bi_max_vecs allows, so we have to ALIGN_DOWN the
+	 * result to ensure the bio's total size is correct. The remainder of
+	 * the iov data will be picked up in the next bio iteration.
+	 */
 	size = iov_iter_get_pages(iter, pages, LONG_MAX, nr_pages, &offset);
+	if (size > 0)
+		size = ALIGN_DOWN(size, bdev_logical_block_size(bio->bi_bdev));
 	if (unlikely(size <= 0))
 		return size ? size : -EFAULT;
 
diff --git a/block/fops.c b/block/fops.c
index 9d32df6fc315..86d3cab9bf93 100644
--- a/block/fops.c
+++ b/block/fops.c
@@ -45,8 +45,8 @@ static unsigned int dio_bio_write_op(struct kiocb *iocb)
 static bool blkdev_dio_unaligned(struct block_device *bdev, loff_t pos,
 			      struct iov_iter *iter)
 {
-	return ((pos | iov_iter_alignment(iter)) &
-	    (bdev_logical_block_size(bdev) - 1));
+	return pos & (bdev_logical_block_size(bdev) - 1) ||
+		!bdev_iter_is_aligned(bdev, iter);
 }
 
 #define DIO_INLINE_BIO_VECS 4
diff --git a/include/linux/blkdev.h b/include/linux/blkdev.h
index fb5c177708d5..914c613d81da 100644
--- a/include/linux/blkdev.h
+++ b/include/linux/blkdev.h
@@ -425,6 +425,11 @@ struct request_queue {
 	unsigned long		nr_requests;	/* Max # of requests */
 
 	unsigned int		dma_pad_mask;
+	/*
+	 * Drivers that set dma_alignment to less than 511 must be prepared to
+	 * handle individual bvec's that are not a multiple of a SECTOR_SIZE
+	 * due to possible offsets.
+	 */
 	unsigned int		dma_alignment;
 
 #ifdef CONFIG_BLK_INLINE_ENCRYPTION
-- 
2.30.2



  parent reply	other threads:[~2022-06-10 20:04 UTC|newest]

Thread overview: 42+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-10 19:58 [PATCHv6 00/11] direct-io dma alignment Keith Busch
2022-06-10 19:58 ` [PATCHv6 01/11] block: fix infinite loop for invalid zone append Keith Busch
2022-06-10 19:58 ` [PATCHv6 02/11] block/bio: remove duplicate append pages code Keith Busch
2022-06-10 19:58 ` [PATCHv6 03/11] block: export dma_alignment attribute Keith Busch
2022-06-10 19:58 ` [PATCHv6 04/11] block: introduce bdev_dma_alignment helper Keith Busch
2022-06-10 19:58 ` [PATCHv6 05/11] block: add a helper function for dio alignment Keith Busch
2022-07-22 21:53   ` Bart Van Assche
2022-06-10 19:58 ` [PATCHv6 06/11] block/merge: count bytes instead of sectors Keith Busch
2022-07-22 21:57   ` Bart Van Assche
2022-06-10 19:58 ` [PATCHv6 07/11] block/bounce: " Keith Busch
2022-06-13 14:22   ` Christoph Hellwig
2022-07-22 22:01   ` Bart Van Assche
2022-07-25 14:46     ` Keith Busch
2022-06-10 19:58 ` [PATCHv6 08/11] iov: introduce iov_iter_aligned Keith Busch
2022-06-10 19:58 ` [PATCHv6 09/11] block: introduce bdev_iter_is_aligned helper Keith Busch
2022-06-10 19:58 ` Keith Busch [this message]
2022-06-10 19:58 ` [PATCHv6 11/11] iomap: add support for dma aligned direct-io Keith Busch
2022-06-23 18:29   ` Eric Farman
2022-06-23 18:51     ` Keith Busch
2022-06-23 19:11       ` Keith Busch
2022-06-23 20:32         ` Eric Farman
2022-06-23 21:34           ` Eric Farman
2022-06-27 15:21             ` Eric Farman
2022-06-27 15:36               ` Keith Busch
2022-06-28  9:00                 ` Halil Pasic
2022-06-28 15:20                   ` Eric Farman
2022-06-29  3:18                     ` Eric Farman
2022-06-29  3:52                       ` Keith Busch
2022-06-29 18:04                         ` Eric Farman
2022-06-29 19:07                           ` Keith Busch
2022-06-29 19:28                             ` Eric Farman
2022-06-30  5:45                             ` Christian Borntraeger
2022-07-22  7:36   ` Eric Biggers
2022-07-22 14:43     ` Keith Busch
2022-07-22 18:01       ` Eric Biggers
2022-07-22 20:26         ` Keith Busch
2022-07-25 18:19           ` Eric Biggers
2022-07-24  2:13         ` Jaegeuk Kim
2022-07-22 17:53     ` Darrick J. Wong
2022-07-22 18:12       ` Eric Biggers
2022-07-23  5:03         ` Darrick J. Wong
2022-06-13 21:22 ` [PATCHv6 00/11] direct-io dma alignment Jens Axboe

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=20220610195830.3574005-11-kbusch@fb.com \
    --to=kbusch@fb.com \
    --cc=Kernel-team@fb.com \
    --cc=axboe@kernel.dk \
    --cc=bvanassche@acm.org \
    --cc=damien.lemoal@opensource.wdc.com \
    --cc=ebiggers@kernel.org \
    --cc=hch@lst.de \
    --cc=kbusch@kernel.org \
    --cc=linux-block@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-nvme@lists.infradead.org \
    --cc=pankydev8@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