All of lore.kernel.org
 help / color / mirror / Atom feed
From: Dave Chinner <david@fromorbit.com>
To: linux-xfs@vger.kernel.org
Subject: [PATCH 3/3] xfs: alignment check bio buffers
Date: Wed, 21 Aug 2019 18:38:20 +1000	[thread overview]
Message-ID: <20190821083820.11725-4-david@fromorbit.com> (raw)
In-Reply-To: <20190821083820.11725-1-david@fromorbit.com>

From: Dave Chinner <dchinner@redhat.com>

Add memory buffer alignment validation checks to bios built in XFS
to catch bugs that will result in silent data corruption in block
drivers that cannot handle unaligned memory buffers but don't
validate the incoming buffer alignment is correct.

Known drivers with these issues are xenblk, brd and pmem.

Despite there being nothing XFS specific to xfs_bio_add_page(), this
function was created to do the required validation because the block
layer developers that keep telling us that is not possible to
validate buffer alignment in bio_add_page(), and even if it was
possible it would be too much overhead to do at runtime.

Signed-off-by: Dave Chinner <dchinner@redhat.com>
---
 fs/xfs/xfs_bio_io.c | 32 +++++++++++++++++++++++++++++---
 fs/xfs/xfs_buf.c    |  2 +-
 fs/xfs/xfs_linux.h  |  3 +++
 fs/xfs/xfs_log.c    |  6 +++++-
 4 files changed, 38 insertions(+), 5 deletions(-)

diff --git a/fs/xfs/xfs_bio_io.c b/fs/xfs/xfs_bio_io.c
index e2148f2d5d6b..fbaea643c000 100644
--- a/fs/xfs/xfs_bio_io.c
+++ b/fs/xfs/xfs_bio_io.c
@@ -9,6 +9,27 @@ static inline unsigned int bio_max_vecs(unsigned int count)
 	return min_t(unsigned, howmany(count, PAGE_SIZE), BIO_MAX_PAGES);
 }
 
+int
+xfs_bio_add_page(
+	struct bio	*bio,
+	struct page	*page,
+	unsigned int	len,
+	unsigned int	offset)
+{
+	struct request_queue	*q = bio->bi_disk->queue;
+	bool		same_page = false;
+
+	if (WARN_ON_ONCE(!blk_rq_aligned(q, len, offset)))
+		return -EIO;
+
+	if (!__bio_try_merge_page(bio, page, len, offset, &same_page)) {
+		if (bio_full(bio, len))
+			return 0;
+		__bio_add_page(bio, page, len, offset);
+	}
+	return len;
+}
+
 int
 xfs_rw_bdev(
 	struct block_device	*bdev,
@@ -20,7 +41,7 @@ xfs_rw_bdev(
 {
 	unsigned int		is_vmalloc = is_vmalloc_addr(data);
 	unsigned int		left = count;
-	int			error;
+	int			error, ret = 0;
 	struct bio		*bio;
 
 	if (is_vmalloc && op == REQ_OP_WRITE)
@@ -36,9 +57,12 @@ xfs_rw_bdev(
 		unsigned int	off = offset_in_page(data);
 		unsigned int	len = min_t(unsigned, left, PAGE_SIZE - off);
 
-		while (bio_add_page(bio, page, len, off) != len) {
+		while ((ret = xfs_bio_add_page(bio, page, len, off)) != len) {
 			struct bio	*prev = bio;
 
+			if (ret < 0)
+				goto submit;
+
 			bio = bio_alloc(GFP_KERNEL, bio_max_vecs(left));
 			bio_copy_dev(bio, prev);
 			bio->bi_iter.bi_sector = bio_end_sector(prev);
@@ -48,14 +72,16 @@ xfs_rw_bdev(
 			submit_bio(prev);
 		}
 
+		ret = 0;
 		data += len;
 		left -= len;
 	} while (left > 0);
 
+submit:
 	error = submit_bio_wait(bio);
 	bio_put(bio);
 
 	if (is_vmalloc && op == REQ_OP_READ)
 		invalidate_kernel_vmap_range(data, count);
-	return error;
+	return ret ? ret : error;
 }
diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c
index 7bd1f31febfc..a2d499baee9c 100644
--- a/fs/xfs/xfs_buf.c
+++ b/fs/xfs/xfs_buf.c
@@ -1294,7 +1294,7 @@ xfs_buf_ioapply_map(
 		if (nbytes > size)
 			nbytes = size;
 
-		rbytes = bio_add_page(bio, bp->b_pages[page_index], nbytes,
+		rbytes = xfs_bio_add_page(bio, bp->b_pages[page_index], nbytes,
 				      offset);
 		if (rbytes < nbytes)
 			break;
diff --git a/fs/xfs/xfs_linux.h b/fs/xfs/xfs_linux.h
index ca15105681ca..e71c7bf3e714 100644
--- a/fs/xfs/xfs_linux.h
+++ b/fs/xfs/xfs_linux.h
@@ -145,6 +145,9 @@ static inline void delay(long ticks)
 	schedule_timeout_uninterruptible(ticks);
 }
 
+int xfs_bio_add_page(struct bio *bio, struct page *page, unsigned int len,
+			unsigned int offset);
+
 /*
  * XFS wrapper structure for sysfs support. It depends on external data
  * structures and is embedded in various internal data structures to implement
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index 1830d185d7fc..52f7d840d09e 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -1673,8 +1673,12 @@ xlog_map_iclog_data(
 		struct page	*page = kmem_to_page(data);
 		unsigned int	off = offset_in_page(data);
 		size_t		len = min_t(size_t, count, PAGE_SIZE - off);
+		int		ret;
 
-		WARN_ON_ONCE(bio_add_page(bio, page, len, off) != len);
+		ret = xfs_bio_add_page(bio, page, len, off);
+		WARN_ON_ONCE(ret != len);
+		if (ret < 0)
+			break;
 
 		data += len;
 		count -= len;
-- 
2.23.0.rc1


  parent reply	other threads:[~2019-08-21  8:38 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-21  8:38 [PATCH 0/3] xfs: avoid IO issues unaligned memory allocation Dave Chinner
2019-08-21  8:38 ` [PATCH 1/3] xfs: add kmem allocation trace points Dave Chinner
2019-08-21 13:34   ` Brian Foster
2019-08-21 23:20   ` Christoph Hellwig
2019-08-21  8:38 ` [PATCH 2/3] xfs: add kmem_alloc_io() Dave Chinner
2019-08-21 13:35   ` Brian Foster
2019-08-21 15:08     ` Darrick J. Wong
2019-08-21 21:24       ` Dave Chinner
2019-08-21 15:23     ` Eric Sandeen
2019-08-21 21:14     ` Dave Chinner
2019-08-22 13:40       ` Brian Foster
2019-08-22 22:39         ` Dave Chinner
2019-08-23 12:10           ` Brian Foster
2019-08-21 23:24   ` Christoph Hellwig
2019-08-22  0:31     ` Dave Chinner
2019-08-22  7:59       ` Christoph Hellwig
2019-08-22  8:51         ` Peter Zijlstra
2019-08-22  9:10           ` Peter Zijlstra
2019-08-22 10:14             ` Dave Chinner
2019-08-22 11:14               ` Vlastimil Babka
2019-08-22 12:07                 ` Dave Chinner
2019-08-22 12:19                   ` Vlastimil Babka
2019-08-22 13:17                     ` Dave Chinner
2019-08-22 14:26                       ` Vlastimil Babka
2019-08-26 12:21                         ` Michal Hocko
2019-08-21  8:38 ` Dave Chinner [this message]
2019-08-21 13:39   ` [PATCH 3/3] xfs: alignment check bio buffers Brian Foster
2019-08-21 21:39     ` Dave Chinner
2019-08-22 13:47       ` Brian Foster
2019-08-22 23:03         ` Dave Chinner
2019-08-23 12:33           ` Brian Foster
2019-08-21 23:30     ` Christoph Hellwig
2019-08-22  0:44       ` Dave Chinner
2019-08-21 23:29   ` Christoph Hellwig
2019-08-22  0:37     ` Dave Chinner
2019-08-22  8:03       ` Christoph Hellwig
2019-08-22 10:17         ` Dave Chinner
2019-08-22  2:50     ` Ming Lei
2019-08-22  4:49       ` Dave Chinner
2019-08-22  7:23         ` Ming Lei
2019-08-22  8:08         ` Christoph Hellwig
2019-08-22 10:20           ` Ming Lei
2019-08-23  0:14             ` Christoph Hellwig
2019-08-23  1:19               ` Ming Lei

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=20190821083820.11725-4-david@fromorbit.com \
    --to=david@fromorbit.com \
    --cc=linux-xfs@vger.kernel.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.