From: Christoph Hellwig <hch@infradead.org>
To: Jan Kara <jack@suse.cz>
Cc: Christoph Hellwig <hch@infradead.org>,
Boaz Harrosh <boazh@netapp.com>,
linux-nvdimm@lists.01.org, linux-xfs@vger.kernel.org,
Andy Lutomirski <luto@kernel.org>,
linux-fsdevel@vger.kernel.org, linux-ext4@vger.kernel.org
Subject: Re: [PATCH 13/13] ext4: Support for synchronous DAX faults
Date: Thu, 24 Aug 2017 05:34:52 -0700 [thread overview]
Message-ID: <20170824123452.GB21282@infradead.org> (raw)
In-Reply-To: <20170824123126.GA21282@infradead.org>
On Thu, Aug 24, 2017 at 05:31:26AM -0700, Christoph Hellwig wrote:
> On Thu, Aug 17, 2017 at 06:08:15PM +0200, Jan Kara wrote:
> > We return IOMAP_F_NEEDDSYNC flag from ext4_iomap_begin() for a
> > synchronous write fault when inode has some uncommitted metadata
> > changes. In the fault handler ext4_dax_fault() we then detect this case,
> > call vfs_fsync_range() to make sure all metadata is committed, and call
> > dax_pfn_mkwrite() to mark PTE as writeable. Note that this will also
> > dirty corresponding radix tree entry which is what we want - fsync(2)
> > will still provide data integrity guarantees for applications not using
> > userspace flushing. And applications using userspace flushing can avoid
> > calling fsync(2) and thus avoid the performance overhead.
>
> Why is this only wiered up for the huge_fault handler and not the
> regular?
Ah, turns out ext4 implements ->fault in terms of ->huge_fault.
We'll really need to sort out this mess of fault handlers before
doing too much surgery here..
_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm
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@lists.01.org,
Andy Lutomirski <luto@kernel.org>,
linux-ext4@vger.kernel.org, linux-xfs@vger.kernel.org,
Christoph Hellwig <hch@infradead.org>,
Ross Zwisler <ross.zwisler@linux.intel.com>,
Dan Williams <dan.j.williams@intel.com>,
Boaz Harrosh <boazh@netapp.com>
Subject: Re: [PATCH 13/13] ext4: Support for synchronous DAX faults
Date: Thu, 24 Aug 2017 05:34:52 -0700 [thread overview]
Message-ID: <20170824123452.GB21282@infradead.org> (raw)
In-Reply-To: <20170824123126.GA21282@infradead.org>
On Thu, Aug 24, 2017 at 05:31:26AM -0700, Christoph Hellwig wrote:
> On Thu, Aug 17, 2017 at 06:08:15PM +0200, Jan Kara wrote:
> > We return IOMAP_F_NEEDDSYNC flag from ext4_iomap_begin() for a
> > synchronous write fault when inode has some uncommitted metadata
> > changes. In the fault handler ext4_dax_fault() we then detect this case,
> > call vfs_fsync_range() to make sure all metadata is committed, and call
> > dax_pfn_mkwrite() to mark PTE as writeable. Note that this will also
> > dirty corresponding radix tree entry which is what we want - fsync(2)
> > will still provide data integrity guarantees for applications not using
> > userspace flushing. And applications using userspace flushing can avoid
> > calling fsync(2) and thus avoid the performance overhead.
>
> Why is this only wiered up for the huge_fault handler and not the
> regular?
Ah, turns out ext4 implements ->fault in terms of ->huge_fault.
We'll really need to sort out this mess of fault handlers before
doing too much surgery here..
WARNING: multiple messages have this Message-ID (diff)
From: Christoph Hellwig <hch-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
To: Jan Kara <jack-AlSwsSmVLrQ@public.gmane.org>
Cc: Christoph Hellwig <hch-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>,
Boaz Harrosh <boazh-HgOvQuBEEgTQT0dZR+AlfA@public.gmane.org>,
linux-nvdimm-hn68Rpc1hR1g9hUCZPvPmw@public.gmane.org,
linux-xfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
Andy Lutomirski <luto-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>,
linux-fsdevel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-ext4-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
Subject: Re: [PATCH 13/13] ext4: Support for synchronous DAX faults
Date: Thu, 24 Aug 2017 05:34:52 -0700 [thread overview]
Message-ID: <20170824123452.GB21282@infradead.org> (raw)
In-Reply-To: <20170824123126.GA21282-wEGCiKHe2LqWVfeAwA7xHQ@public.gmane.org>
On Thu, Aug 24, 2017 at 05:31:26AM -0700, Christoph Hellwig wrote:
> On Thu, Aug 17, 2017 at 06:08:15PM +0200, Jan Kara wrote:
> > We return IOMAP_F_NEEDDSYNC flag from ext4_iomap_begin() for a
> > synchronous write fault when inode has some uncommitted metadata
> > changes. In the fault handler ext4_dax_fault() we then detect this case,
> > call vfs_fsync_range() to make sure all metadata is committed, and call
> > dax_pfn_mkwrite() to mark PTE as writeable. Note that this will also
> > dirty corresponding radix tree entry which is what we want - fsync(2)
> > will still provide data integrity guarantees for applications not using
> > userspace flushing. And applications using userspace flushing can avoid
> > calling fsync(2) and thus avoid the performance overhead.
>
> Why is this only wiered up for the huge_fault handler and not the
> regular?
Ah, turns out ext4 implements ->fault in terms of ->huge_fault.
We'll really need to sort out this mess of fault handlers before
doing too much surgery here..
next prev parent reply other threads:[~2017-08-24 12:32 UTC|newest]
Thread overview: 142+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-08-17 16:08 [RFC PATCH 0/13 v2] dax, ext4: Synchronous page faults Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` [PATCH 01/13] mm: Remove VM_FAULT_HWPOISON_LARGE_MASK Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` [PATCH 02/13] dax: Simplify arguments of dax_insert_mapping() Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` [PATCH 03/13] dax: Factor out getting of pfn out of iomap Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-18 22:06 ` Ross Zwisler
2017-08-18 22:06 ` Ross Zwisler
2017-08-23 18:30 ` Christoph Hellwig
2017-08-23 18:30 ` Christoph Hellwig
2017-08-17 16:08 ` [PATCH 04/13] dax: Create local variable for VMA in dax_iomap_pte_fault() Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-18 22:08 ` Ross Zwisler
2017-08-18 22:08 ` Ross Zwisler
2017-08-23 18:30 ` Christoph Hellwig
2017-08-23 18:30 ` Christoph Hellwig
2017-08-17 16:08 ` [PATCH 05/13] dax: Create local variable for vmf->flags & FAULT_FLAG_WRITE test Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-18 22:08 ` Ross Zwisler
2017-08-18 22:08 ` Ross Zwisler
2017-08-18 22:08 ` Ross Zwisler
2017-08-23 18:31 ` Christoph Hellwig
2017-08-23 18:31 ` Christoph Hellwig
2017-08-23 18:31 ` Christoph Hellwig
2017-08-17 16:08 ` [PATCH 06/13] dax: Inline dax_insert_mapping() into the callsite Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-18 22:10 ` Ross Zwisler
2017-08-18 22:10 ` Ross Zwisler
2017-08-18 22:10 ` Ross Zwisler
2017-08-23 18:31 ` Christoph Hellwig
2017-08-23 18:31 ` Christoph Hellwig
2017-08-23 18:31 ` Christoph Hellwig
2017-08-17 16:08 ` [PATCH 07/13] dax: Inline dax_pmd_insert_mapping() " Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-18 22:12 ` Ross Zwisler
2017-08-18 22:12 ` Ross Zwisler
2017-08-18 22:12 ` Ross Zwisler
2017-08-23 18:32 ` Christoph Hellwig
2017-08-23 18:32 ` Christoph Hellwig
2017-08-17 16:08 ` [PATCH 08/13] dax: Fix comment describing dax_iomap_fault() Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-18 22:12 ` Ross Zwisler
2017-08-18 22:12 ` Ross Zwisler
2017-08-23 18:32 ` Christoph Hellwig
2017-08-23 18:32 ` Christoph Hellwig
2017-08-17 16:08 ` [PATCH 09/13] dax: Allow dax_iomap_fault() to return pfn Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-21 18:45 ` Ross Zwisler
2017-08-21 18:45 ` Ross Zwisler
2017-08-23 18:34 ` Christoph Hellwig
2017-08-23 18:34 ` Christoph Hellwig
2017-08-23 18:34 ` Christoph Hellwig
2017-08-24 7:26 ` Jan Kara
2017-08-24 7:26 ` Jan Kara
2017-08-17 16:08 ` [PATCH 10/13] mm: Wire up MAP_SYNC Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-21 21:37 ` Ross Zwisler
2017-08-21 21:37 ` Ross Zwisler
2017-08-22 9:36 ` Jan Kara
2017-08-22 9:36 ` Jan Kara
2017-08-21 21:57 ` Ross Zwisler
2017-08-21 21:57 ` Ross Zwisler
2017-08-21 21:57 ` Ross Zwisler
2017-08-22 9:34 ` Jan Kara
2017-08-22 9:34 ` Jan Kara
2017-08-22 17:27 ` Dan Williams
2017-08-22 17:27 ` Dan Williams
2017-08-22 17:27 ` Dan Williams
2017-08-23 18:43 ` Christoph Hellwig
2017-08-23 18:43 ` Christoph Hellwig
2017-08-23 18:43 ` Christoph Hellwig
2017-08-24 7:16 ` Jan Kara
2017-08-24 7:16 ` Jan Kara
2017-08-17 16:08 ` [PATCH 11/13] dax, iomap: Add support for synchronous faults Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-21 18:58 ` Ross Zwisler
2017-08-21 18:58 ` Ross Zwisler
2017-08-22 9:46 ` Jan Kara
2017-08-22 9:46 ` Jan Kara
2017-08-21 21:09 ` Ross Zwisler
2017-08-21 21:09 ` Ross Zwisler
2017-08-22 10:08 ` Jan Kara
2017-08-22 10:08 ` Jan Kara
2017-08-22 10:08 ` Jan Kara
2017-08-24 12:27 ` Christoph Hellwig
2017-08-24 12:27 ` Christoph Hellwig
2017-08-24 12:34 ` Jan Kara
2017-08-24 12:34 ` Jan Kara
2017-08-24 13:38 ` Christoph Hellwig
2017-08-24 13:38 ` Christoph Hellwig
2017-08-24 16:45 ` Jan Kara
2017-08-24 16:45 ` Jan Kara
2017-08-17 16:08 ` [PATCH 12/13] dax: Implement dax_insert_pfn_mkwrite() Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-21 19:01 ` Ross Zwisler
2017-08-21 19:01 ` Ross Zwisler
2017-08-17 16:08 ` [PATCH 13/13] ext4: Support for synchronous DAX faults Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-17 16:08 ` Jan Kara
2017-08-21 19:19 ` Ross Zwisler
2017-08-21 19:19 ` Ross Zwisler
2017-08-22 10:18 ` Jan Kara
2017-08-22 10:18 ` Jan Kara
2017-08-22 10:18 ` Jan Kara
2017-08-23 18:37 ` Christoph Hellwig
2017-08-23 18:37 ` Christoph Hellwig
2017-08-24 7:18 ` Jan Kara
2017-08-24 7:18 ` Jan Kara
2017-08-24 12:31 ` Christoph Hellwig
2017-08-24 12:31 ` Christoph Hellwig
2017-08-24 12:34 ` Christoph Hellwig [this message]
2017-08-24 12:34 ` Christoph Hellwig
2017-08-24 12:34 ` Christoph Hellwig
2017-08-24 12:36 ` Jan Kara
2017-08-24 12:36 ` 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=20170824123452.GB21282@infradead.org \
--to=hch@infradead.org \
--cc=boazh@netapp.com \
--cc=jack@suse.cz \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-nvdimm@lists.01.org \
--cc=linux-xfs@vger.kernel.org \
--cc=luto@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.