From: Christoph Hellwig <hch-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
To: Jan Kara <jack-AlSwsSmVLrQ@public.gmane.org>
Cc: linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-nvdimm-y27Ovi1pjclAfugRpC6u6w@public.gmane.org
Subject: Re: [PATCH 6/6] dax: Avoid page invalidation races and unnecessary radix tree traversals
Date: Fri, 30 Sep 2016 01:55:44 -0700 [thread overview]
Message-ID: <20160930085544.GD22381@infradead.org> (raw)
In-Reply-To: <1474994615-29553-7-git-send-email-jack-AlSwsSmVLrQ@public.gmane.org>
On Tue, Sep 27, 2016 at 06:43:35PM +0200, Jan Kara wrote:
> Currently each filesystem (possibly through generic_file_direct_write()
> or iomap_dax_rw()) takes care of invalidating page tables and evicting
> hole pages from the radix tree when write(2) to the file happens. This
> invalidation is only necessary when there is some block allocation
> resulting from write(2). Furthermore in current place the invalidation
> is racy wrt page fault instantiating a hole page just after we have
> invalidated it.
>
> So perform the page invalidation inside dax_do_io() where we can do it
> only when really necessary and after blocks have been allocated so
> nobody will be instantiating new hole pages anymore.
>
> Signed-off-by: Jan Kara <jack-AlSwsSmVLrQ@public.gmane.org>
This looks fine with the comment from Dave addressed:
Reviewed-by: Christoph Hellwig <hch-jcswGhMUV9g@public.gmane.org>
> + if (buffer_new(bh) &&
> + inode->i_mapping->nrpages) {
Btw, it would be nice if the nrpages check could move into
invalidate_inode_pages2_range instead of having to bother with it in
the callers.
WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch@infradead.org>
To: Jan Kara <jack@suse.cz>
Cc: linux-fsdevel@vger.kernel.org, linux-nvdimm@ml01.01.org,
Dan Williams <dan.j.williams@intel.com>,
Ross Zwisler <ross.zwisler@linux.intel.com>
Subject: Re: [PATCH 6/6] dax: Avoid page invalidation races and unnecessary radix tree traversals
Date: Fri, 30 Sep 2016 01:55:44 -0700 [thread overview]
Message-ID: <20160930085544.GD22381@infradead.org> (raw)
In-Reply-To: <1474994615-29553-7-git-send-email-jack@suse.cz>
On Tue, Sep 27, 2016 at 06:43:35PM +0200, Jan Kara wrote:
> Currently each filesystem (possibly through generic_file_direct_write()
> or iomap_dax_rw()) takes care of invalidating page tables and evicting
> hole pages from the radix tree when write(2) to the file happens. This
> invalidation is only necessary when there is some block allocation
> resulting from write(2). Furthermore in current place the invalidation
> is racy wrt page fault instantiating a hole page just after we have
> invalidated it.
>
> So perform the page invalidation inside dax_do_io() where we can do it
> only when really necessary and after blocks have been allocated so
> nobody will be instantiating new hole pages anymore.
>
> Signed-off-by: Jan Kara <jack@suse.cz>
This looks fine with the comment from Dave addressed:
Reviewed-by: Christoph Hellwig <hch@lst.de>
> + if (buffer_new(bh) &&
> + inode->i_mapping->nrpages) {
Btw, it would be nice if the nrpages check could move into
invalidate_inode_pages2_range instead of having to bother with it in
the callers.
next prev parent reply other threads:[~2016-09-30 8:55 UTC|newest]
Thread overview: 51+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-27 16:43 [PATCH 0/6] dax: Page invalidation fixes Jan Kara
2016-09-27 16:43 ` Jan Kara
2016-09-27 16:43 ` [PATCH 1/6] dax: Do not warn about BH_New buffers Jan Kara
2016-09-27 16:43 ` Jan Kara
[not found] ` <1474994615-29553-2-git-send-email-jack-AlSwsSmVLrQ@public.gmane.org>
2016-09-30 8:53 ` Christoph Hellwig
2016-09-30 8:53 ` Christoph Hellwig
2016-09-27 16:43 ` [PATCH 2/6] ext2: Return BH_New buffers for zeroed blocks Jan Kara
2016-09-27 16:43 ` Jan Kara
[not found] ` <1474994615-29553-3-git-send-email-jack-AlSwsSmVLrQ@public.gmane.org>
2016-09-30 8:53 ` Christoph Hellwig
2016-09-30 8:53 ` Christoph Hellwig
2016-09-27 16:43 ` [PATCH 3/6] ext4: Remove clearing of BH_New bit " Jan Kara
2016-09-27 16:43 ` Jan Kara
[not found] ` <1474994615-29553-4-git-send-email-jack-AlSwsSmVLrQ@public.gmane.org>
2016-09-30 8:53 ` Christoph Hellwig
2016-09-30 8:53 ` Christoph Hellwig
2016-09-27 16:43 ` [PATCH 4/6] xfs: Set BH_New for allocated DAX blocks in __xfs_get_blocks() Jan Kara
2016-09-27 16:43 ` Jan Kara
[not found] ` <1474994615-29553-5-git-send-email-jack-AlSwsSmVLrQ@public.gmane.org>
2016-09-27 17:01 ` Christoph Hellwig
2016-09-27 17:01 ` Christoph Hellwig
2016-09-27 17:17 ` Jan Kara
2016-09-28 0:22 ` Dave Chinner
2016-10-03 14:44 ` Jan Kara
2016-10-03 14:44 ` Jan Kara
[not found] ` <20160927171707.GA30327-4I4JzKEfoa/jFM9bn6wA6Q@public.gmane.org>
2016-09-28 2:13 ` Christoph Hellwig
2016-09-28 2:13 ` Christoph Hellwig
2016-10-03 14:42 ` Jan Kara
[not found] ` <20161003144246.GA14183-4I4JzKEfoa/jFM9bn6wA6Q@public.gmane.org>
2016-10-03 16:39 ` Christoph Hellwig
2016-10-03 16:39 ` Christoph Hellwig
2016-09-27 16:43 ` [PATCH 5/6] mm: Invalidate DAX radix tree entries only if appropriate Jan Kara
2016-09-27 16:43 ` Jan Kara
2016-09-28 0:18 ` Dave Chinner
2016-09-28 0:18 ` Dave Chinner
2016-10-03 14:52 ` Jan Kara
2016-10-03 14:52 ` Jan Kara
[not found] ` <1474994615-29553-6-git-send-email-jack-AlSwsSmVLrQ@public.gmane.org>
2016-09-30 8:57 ` Christoph Hellwig
2016-09-30 8:57 ` Christoph Hellwig
2016-10-03 12:56 ` Jan Kara
2016-09-27 16:43 ` [PATCH 6/6] dax: Avoid page invalidation races and unnecessary radix tree traversals Jan Kara
2016-09-27 16:43 ` Jan Kara
2016-09-28 0:20 ` Dave Chinner
2016-09-28 0:20 ` Dave Chinner
2016-10-03 14:58 ` Jan Kara
2016-10-03 14:58 ` Jan Kara
2016-10-03 23:40 ` Dave Chinner
2016-10-03 23:40 ` Dave Chinner
[not found] ` <1474994615-29553-7-git-send-email-jack-AlSwsSmVLrQ@public.gmane.org>
2016-09-30 8:55 ` Christoph Hellwig [this message]
2016-09-30 8:55 ` Christoph Hellwig
[not found] ` <20160930085544.GD22381-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
2016-10-03 15:02 ` Jan Kara
2016-10-03 15:02 ` Jan Kara
[not found] ` <1474994615-29553-1-git-send-email-jack-AlSwsSmVLrQ@public.gmane.org>
2016-09-30 8:59 ` [PATCH 0/6] dax: Page invalidation fixes Christoph Hellwig
2016-09-30 8:59 ` Christoph Hellwig
2016-10-03 13:01 ` 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=20160930085544.GD22381@infradead.org \
--to=hch-wegcikhe2lqwvfeawa7xhq@public.gmane.org \
--cc=jack-AlSwsSmVLrQ@public.gmane.org \
--cc=linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=linux-nvdimm-y27Ovi1pjclAfugRpC6u6w@public.gmane.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.