From: Jan Kara <jack@suse.cz>
To: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>,
stable@vger.kernel.org,
Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>,
Ross Zwisler <ross.zwisler@linux.intel.com>,
Andrew Morton <akpm@linux-foundation.org>,
Jan Kara <jack@suse.cz>, Jens Axboe <axboe@kernel.dk>,
Johannes Weiner <hannes@cmpxchg.org>,
Alexey Kuznetsov <kuznet@virtuozzo.com>,
Christoph Hellwig <hch@lst.de>,
Nikolay Borisov <n.borisov.lkml@gmail.com>,
linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
linux-mm@kvack.org
Subject: Re: [PATCH v2 1/4] fs: fix data invalidation in the cleancache during direct IO
Date: Tue, 25 Apr 2017 10:25:45 +0200 [thread overview]
Message-ID: <20170425082545.GB2793@quack2.suse.cz> (raw)
In-Reply-To: <20170424164135.22350-2-aryabinin@virtuozzo.com>
On Mon 24-04-17 19:41:32, Andrey Ryabinin wrote:
> Some direct IO write fs hooks call invalidate_inode_pages2[_range]()
> conditionally iff mapping->nrpages is not zero. This can't be right,
> because invalidate_inode_pages2[_ragne]() also invalidate data in
> the cleancache via cleancache_invalidate_inode() call.
> So if page cache is empty but there is some data in the cleancache,
> buffered read after direct IO write would get stale data from
> the cleancache.
>
> Also it doesn't feel right to check only for ->nrpages because
> invalidate_inode_pages2[_range] invalidates exceptional entries as well.
>
> Fix this by calling invalidate_inode_pages2[_range]() regardless of nrpages
> state.
>
> Note: nfs,cifs,9p doesn't need similar fix because the never call
> cleancache_get_page() (nor directly, nor via mpage_readpage[s]()), so they
> are not affected by this bug.
>
> Fixes: c515e1fd361c ("mm/fs: add hooks to support cleancache")
> Signed-off-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
> Acked-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
> Cc: <stable@vger.kernel.org>
OK, looks good. You can add:
Reviewed-by: Jan Kara <jack@suse.cz>
Honza
> ---
> fs/iomap.c | 18 ++++++++----------
> mm/filemap.c | 26 +++++++++++---------------
> 2 files changed, 19 insertions(+), 25 deletions(-)
>
> diff --git a/fs/iomap.c b/fs/iomap.c
> index cdeed39..f6a6013 100644
> --- a/fs/iomap.c
> +++ b/fs/iomap.c
> @@ -881,16 +881,14 @@ iomap_dio_rw(struct kiocb *iocb, struct iov_iter *iter,
> flags |= IOMAP_WRITE;
> }
>
> - if (mapping->nrpages) {
> - ret = filemap_write_and_wait_range(mapping, start, end);
> - if (ret)
> - goto out_free_dio;
> + ret = filemap_write_and_wait_range(mapping, start, end);
> + if (ret)
> + goto out_free_dio;
>
> - ret = invalidate_inode_pages2_range(mapping,
> - start >> PAGE_SHIFT, end >> PAGE_SHIFT);
> - WARN_ON_ONCE(ret);
> - ret = 0;
> - }
> + ret = invalidate_inode_pages2_range(mapping,
> + start >> PAGE_SHIFT, end >> PAGE_SHIFT);
> + WARN_ON_ONCE(ret);
> + ret = 0;
>
> inode_dio_begin(inode);
>
> @@ -945,7 +943,7 @@ iomap_dio_rw(struct kiocb *iocb, struct iov_iter *iter,
> * one is a pretty crazy thing to do, so we don't support it 100%. If
> * this invalidation fails, tough, the write still worked...
> */
> - if (iov_iter_rw(iter) == WRITE && mapping->nrpages) {
> + if (iov_iter_rw(iter) == WRITE) {
> int err = invalidate_inode_pages2_range(mapping,
> start >> PAGE_SHIFT, end >> PAGE_SHIFT);
> WARN_ON_ONCE(err);
> diff --git a/mm/filemap.c b/mm/filemap.c
> index 9eab40e..b7b973b 100644
> --- a/mm/filemap.c
> +++ b/mm/filemap.c
> @@ -2720,18 +2720,16 @@ generic_file_direct_write(struct kiocb *iocb, struct iov_iter *from)
> * about to write. We do this *before* the write so that we can return
> * without clobbering -EIOCBQUEUED from ->direct_IO().
> */
> - if (mapping->nrpages) {
> - written = invalidate_inode_pages2_range(mapping,
> + written = invalidate_inode_pages2_range(mapping,
> pos >> PAGE_SHIFT, end);
> - /*
> - * If a page can not be invalidated, return 0 to fall back
> - * to buffered write.
> - */
> - if (written) {
> - if (written == -EBUSY)
> - return 0;
> - goto out;
> - }
> + /*
> + * If a page can not be invalidated, return 0 to fall back
> + * to buffered write.
> + */
> + if (written) {
> + if (written == -EBUSY)
> + return 0;
> + goto out;
> }
>
> written = mapping->a_ops->direct_IO(iocb, from);
> @@ -2744,10 +2742,8 @@ generic_file_direct_write(struct kiocb *iocb, struct iov_iter *from)
> * so we don't support it 100%. If this invalidation
> * fails, tough, the write still worked...
> */
> - if (mapping->nrpages) {
> - invalidate_inode_pages2_range(mapping,
> - pos >> PAGE_SHIFT, end);
> - }
> + invalidate_inode_pages2_range(mapping,
> + pos >> PAGE_SHIFT, end);
>
> if (written > 0) {
> pos += written;
> --
> 2.10.2
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
next prev parent reply other threads:[~2017-04-25 8:26 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-04-14 14:07 [PATCH 0/4] Properly invalidate data in the cleancache Andrey Ryabinin
2017-04-14 14:07 ` [PATCH 1/4] fs: fix data invalidation in the cleancache during direct IO Andrey Ryabinin
2017-04-18 19:38 ` Ross Zwisler
2017-04-19 15:11 ` Andrey Ryabinin
2017-04-19 19:28 ` Ross Zwisler
2017-04-20 14:35 ` Jan Kara
2017-04-20 14:44 ` Jan Kara
2017-04-20 19:14 ` Ross Zwisler
2017-04-21 3:44 ` [PATCH 1/2] dax: prevent invalidation of mapped DAX entries Ross Zwisler
2017-04-21 3:44 ` [PATCH 2/2] dax: fix data corruption due to stale mmap reads Ross Zwisler
2017-04-25 11:10 ` Jan Kara
2017-04-25 22:59 ` Ross Zwisler
2017-04-26 8:52 ` Jan Kara
2017-04-26 22:52 ` Ross Zwisler
2017-04-27 7:26 ` Jan Kara
2017-05-01 22:38 ` Ross Zwisler
2017-05-04 9:12 ` Jan Kara
2017-05-01 22:59 ` Dan Williams
2017-04-25 10:10 ` [PATCH 1/2] dax: prevent invalidation of mapped DAX entries Jan Kara
2017-05-01 16:54 ` Ross Zwisler
2017-04-18 22:46 ` [PATCH 1/4] fs: fix data invalidation in the cleancache during direct IO Andrew Morton
2017-04-19 15:15 ` Andrey Ryabinin
2017-04-14 14:07 ` [PATCH 2/4] fs/block_dev: always invalidate cleancache in invalidate_bdev() Andrey Ryabinin
2017-04-18 18:51 ` Nikolay Borisov
2017-04-19 13:22 ` Andrey Ryabinin
2017-04-14 14:07 ` [PATCH 3/4] mm/truncate: bail out early from invalidate_inode_pages2_range() if mapping is empty Andrey Ryabinin
2017-04-14 14:07 ` [PATCH 4/4] mm/truncate: avoid pointless cleancache_invalidate_inode() calls Andrey Ryabinin
2017-04-18 15:24 ` [PATCH 0/4] Properly invalidate data in the cleancache Konrad Rzeszutek Wilk
2017-04-24 16:41 ` [PATCH v2 " Andrey Ryabinin
2017-04-24 16:41 ` [PATCH v2 1/4] fs: fix data invalidation in the cleancache during direct IO Andrey Ryabinin
2017-04-25 8:25 ` Jan Kara [this message]
2017-04-24 16:41 ` [PATCH v2 2/4] fs/block_dev: always invalidate cleancache in invalidate_bdev() Andrey Ryabinin
2017-04-25 8:34 ` Jan Kara
2017-04-24 16:41 ` [PATCH v2 3/4] mm/truncate: bail out early from invalidate_inode_pages2_range() if mapping is empty Andrey Ryabinin
2017-04-25 8:37 ` Jan Kara
2017-04-24 16:41 ` [PATCH v2 4/4] mm/truncate: avoid pointless cleancache_invalidate_inode() calls Andrey Ryabinin
2017-04-25 8:41 ` Jan Kara
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=20170425082545.GB2793@quack2.suse.cz \
--to=jack@suse.cz \
--cc=akpm@linux-foundation.org \
--cc=aryabinin@virtuozzo.com \
--cc=axboe@kernel.dk \
--cc=hannes@cmpxchg.org \
--cc=hch@lst.de \
--cc=konrad.wilk@oracle.com \
--cc=kuznet@virtuozzo.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=n.borisov.lkml@gmail.com \
--cc=ross.zwisler@linux.intel.com \
--cc=stable@vger.kernel.org \
--cc=viro@zeniv.linux.org.uk \
/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).