linux-block.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jens Axboe <axboe@kernel.dk>
To: Hannes Reinecke <hare@suse.de>
Cc: Christoph Hellwig <hch@lst.de>,
	"linux-block@vger.kernel.org" <linux-block@vger.kernel.org>,
	Martin Wilck <mwilck@suse.com>
Subject: Re: Silent data corruption in blkdev_direct_IO()
Date: Thu, 12 Jul 2018 09:08:57 -0600	[thread overview]
Message-ID: <3419a3ae-da82-9c20-26e1-7c9ed14ff8ed@kernel.dk> (raw)
In-Reply-To: <e36f28d8-0eb8-1600-f327-8b066eeaf45c@suse.de>

On 7/12/18 8:36 AM, Hannes Reinecke wrote:
> Hi Jens, Christoph,
> 
> we're currently hunting down a silent data corruption occurring due to
> commit 72ecad22d9f1 ("block: support a full bio worth of IO for
> simplified bdev direct-io").
> 
> While the whole thing is still hazy on the details, the one thing we've
> found is that reverting that patch fixes the data corruption.
> 
> And looking closer, I've found this:
> 
> static ssize_t
> blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
> {
> 	int nr_pages;
> 
> 	nr_pages = iov_iter_npages(iter, BIO_MAX_PAGES + 1);
> 	if (!nr_pages)
> 		return 0;
> 	if (is_sync_kiocb(iocb) && nr_pages <= BIO_MAX_PAGES)
> 		return __blkdev_direct_IO_simple(iocb, iter, nr_pages);
> 
> 	return __blkdev_direct_IO(iocb, iter, min(nr_pages, BIO_MAX_PAGES));
> }
> 
> When checking the call path
> __blkdev_direct_IO()->bio_alloc_bioset()->bvec_alloc()
> I found that bvec_alloc() will fail if nr_pages > BIO_MAX_PAGES.
> 
> So why is there the check for 'nr_pages <= BIO_MAX_PAGES' ?
> It's not that we can handle it in __blkdev_direct_IO() ...

The logic could be cleaned up like below, the sync part is really all
we care about. What is the test case for this? async or sync?

I also don't remember why it's BIO_MAX_PAGES + 1...

diff --git a/fs/block_dev.c b/fs/block_dev.c
index 0dd87aaeb39a..14ef3d71b55f 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -424,13 +424,13 @@ blkdev_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 {
 	int nr_pages;
 
-	nr_pages = iov_iter_npages(iter, BIO_MAX_PAGES + 1);
+	nr_pages = iov_iter_npages(iter, BIO_MAX_PAGES);
 	if (!nr_pages)
 		return 0;
-	if (is_sync_kiocb(iocb) && nr_pages <= BIO_MAX_PAGES)
+	if (is_sync_kiocb(iocb))
 		return __blkdev_direct_IO_simple(iocb, iter, nr_pages);
 
-	return __blkdev_direct_IO(iocb, iter, min(nr_pages, BIO_MAX_PAGES));
+	return __blkdev_direct_IO(iocb, iter, nr_pages);
 }
 
 static __init int blkdev_init(void)

-- 
Jens Axboe

  reply	other threads:[~2018-07-12 15:08 UTC|newest]

Thread overview: 54+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-07-12 14:36 Silent data corruption in blkdev_direct_IO() Hannes Reinecke
2018-07-12 15:08 ` Jens Axboe [this message]
2018-07-12 16:11   ` Martin Wilck
2018-07-12 16:14   ` Hannes Reinecke
2018-07-12 16:20     ` Jens Axboe
2018-07-12 16:42       ` Jens Axboe
2018-07-13  6:47         ` Martin Wilck
2018-07-13 16:56         ` Martin Wilck
2018-07-13 18:00           ` Jens Axboe
2018-07-13 18:50             ` Jens Axboe
2018-07-13 22:21               ` Martin Wilck
2018-07-13 20:48             ` Martin Wilck
2018-07-13 20:52               ` Jens Axboe
2018-07-16 19:05                 ` Martin Wilck
2018-07-12 23:29 ` Ming Lei
2018-07-13 18:54   ` Jens Axboe
2018-07-13 22:29     ` Martin Wilck
2018-07-16 11:45       ` Ming Lei
2018-07-18  0:07         ` Martin Wilck
2018-07-18  2:48           ` Ming Lei
2018-07-18  7:32             ` Martin Wilck
2018-07-18  7:54               ` Ming Lei
2018-07-18  9:20                 ` Johannes Thumshirn
2018-07-18 11:40                   ` Jan Kara
2018-07-18 11:57                     ` Jan Kara
2018-07-19  9:39                 ` [PATCH 0/2] Fix silent " Martin Wilck
2018-07-19  9:39                   ` [PATCH 1/2] block: bio_iov_iter_get_pages: fix size of last iovec Martin Wilck
2018-07-19 10:05                     ` Hannes Reinecke
2018-07-19 10:09                     ` Ming Lei
2018-07-19 10:20                     ` Jan Kara
2018-07-19 14:52                     ` Christoph Hellwig
2018-07-19  9:39                   ` [PATCH 2/2] blkdev: __blkdev_direct_IO_simple: make sure to fill up the bio Martin Wilck
2018-07-19 10:06                     ` Hannes Reinecke
2018-07-19 10:21                     ` Ming Lei
2018-07-19 10:37                       ` Jan Kara
2018-07-19 10:46                         ` Ming Lei
2018-07-19 11:08                         ` Al Viro
2018-07-19 14:53                           ` Christoph Hellwig
2018-07-19 15:06                             ` Jan Kara
2018-07-19 15:11                               ` Christoph Hellwig
2018-07-19 19:21                                 ` Martin Wilck
2018-07-19 19:34                             ` Martin Wilck
2018-07-19 10:45                     ` Jan Kara
2018-07-19 12:23                       ` Martin Wilck
2018-07-19 15:15                         ` Jan Kara
2018-07-19 20:01                           ` Martin Wilck
2018-07-19 11:04                     ` Ming Lei
2018-07-19 11:56                       ` Jan Kara
2018-07-19 12:20                         ` Ming Lei
2018-07-19 15:21                           ` Jan Kara
2018-07-19 19:06                             ` Martin Wilck
2018-07-19 12:25                         ` Martin Wilck
2018-07-19 10:08                   ` [PATCH 0/2] Fix silent data corruption in blkdev_direct_IO() Hannes Reinecke
2018-07-19 14:50                   ` Christoph Hellwig

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=3419a3ae-da82-9c20-26e1-7c9ed14ff8ed@kernel.dk \
    --to=axboe@kernel.dk \
    --cc=hare@suse.de \
    --cc=hch@lst.de \
    --cc=linux-block@vger.kernel.org \
    --cc=mwilck@suse.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;
as well as URLs for NNTP newsgroup(s).