From: Matthew Bobrowski <mbobrowski@mbobrowski.org>
To: Christoph Hellwig <hch@infradead.org>
Cc: tytso@mit.edu, jack@suse.cz, adilger.kernel@dilger.ca,
linux-ext4@vger.kernel.org, linux-fsdevel@vger.kernel.org,
david@fromorbit.com, darrick.wong@oracle.com
Subject: Re: [PATCH v3 4/6] ext4: reorder map.m_flags checks in ext4_iomap_begin()
Date: Tue, 17 Sep 2019 22:48:59 +1000 [thread overview]
Message-ID: <20190917124859.GC17286@bobrowski> (raw)
In-Reply-To: <20190916120533.GB4005@infradead.org>
On Mon, Sep 16, 2019 at 05:05:33AM -0700, Christoph Hellwig wrote:
> On Thu, Sep 12, 2019 at 09:04:30PM +1000, Matthew Bobrowski wrote:
> > @@ -3581,10 +3581,21 @@ static int ext4_iomap_begin(struct inode *inode, loff_t offset, loff_t length,
> > iomap->type = delalloc ? IOMAP_DELALLOC : IOMAP_HOLE;
> > iomap->addr = IOMAP_NULL_ADDR;
> > } else {
> > - if (map.m_flags & EXT4_MAP_MAPPED) {
> > - iomap->type = IOMAP_MAPPED;
> > - } else if (map.m_flags & EXT4_MAP_UNWRITTEN) {
> > + /*
> > + * Flags passed to ext4_map_blocks() for direct IO
> > + * writes can result in m_flags having both
> > + * EXT4_MAP_MAPPED and EXT4_MAP_UNWRITTEN bits set. In
> > + * order for allocated unwritten extents to be
> > + * converted to written extents in the end_io handler
> > + * correctly, we need to ensure that the iomap->type
> > + * is also set appropriately in that case. Thus, we
> > + * need to check whether EXT4_MAP_UNWRITTEN is set
> > + * first.
> > + */
> > + if (map.m_flags & EXT4_MAP_UNWRITTEN) {
> > iomap->type = IOMAP_UNWRITTEN;
> > + } else if (map.m_flags & EXT4_MAP_MAPPED) {
> > + iomap->type = IOMAP_MAPPED;
>
> I think much of this would benefit a lot from just being split up.
> I hacked up a patch last week that split the ext4 direct I/O code
> a bit, but this is completely untested and needs further splitup,
> but maybe you can take it as an inspiration for your series?
Nice, I really like this! :-)
The ext4_iomap_begin() callback is kind of already getting larger than it
should have to be and I can only see it growing moving forward, so why not
split it up now.
> E.g. at least one helper for filling out the iomap from the ext4
> map data, and one for the seek unwritten extent reporting. The
> split of the overall iomap ops seemed useful to me, but might not
> be as important with the other cleanups:
Yeah, I think I'll leave the iomap operations as they are for now. Something
to definitely consider at a later point though.
--<M>--
next prev parent reply other threads:[~2019-09-17 12:49 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-12 11:03 [PATCH v3 0/6] ext4: port direct IO to iomap infrastructure Matthew Bobrowski
2019-09-12 11:03 ` [PATCH v3 1/6] ext4: introduce direct IO read path using " Matthew Bobrowski
2019-09-16 12:00 ` Christoph Hellwig
2019-09-16 13:07 ` Matthew Bobrowski
2019-09-12 11:04 ` [PATCH v3 2/6] ext4: move inode extension/truncate code out from ext4_iomap_end() Matthew Bobrowski
2019-09-23 16:21 ` Jan Kara
2019-09-24 9:50 ` Matthew Bobrowski
2019-09-24 13:13 ` Jan Kara
2019-09-12 11:04 ` [PATCH v3 3/6] iomap: split size and error for iomap_dio_rw ->end_io Matthew Bobrowski
2019-09-12 11:04 ` [PATCH v3 4/6] ext4: reorder map.m_flags checks in ext4_iomap_begin() Matthew Bobrowski
2019-09-16 12:05 ` Christoph Hellwig
2019-09-17 12:48 ` Matthew Bobrowski [this message]
2019-09-23 15:08 ` Jan Kara
2019-09-24 9:35 ` Matthew Bobrowski
2019-09-12 11:04 ` [PATCH v3 5/6] ext4: introduce direct IO write path using iomap infrastructure Matthew Bobrowski
2019-09-16 4:37 ` Ritesh Harjani
2019-09-16 10:14 ` Matthew Bobrowski
2019-09-16 12:12 ` Christoph Hellwig
2019-09-16 22:37 ` Matthew Bobrowski
2019-09-17 9:00 ` Ritesh Harjani
2019-09-17 9:02 ` Christoph Hellwig
2019-09-17 10:12 ` Ritesh Harjani
2019-09-17 12:39 ` Matthew Bobrowski
2019-09-24 10:57 ` Jan Kara
2019-09-17 9:06 ` Christoph Hellwig
2019-09-17 11:31 ` Matthew Bobrowski
2019-09-20 13:24 ` Matthew Bobrowski
2019-09-23 21:10 ` Jan Kara
2019-09-24 10:29 ` Matthew Bobrowski
2019-09-24 14:13 ` Jan Kara
2019-09-25 7:14 ` Matthew Bobrowski
2019-09-25 8:40 ` Jan Kara
2019-09-12 11:05 ` [PATCH v3 6/6] ext4: cleanup legacy buffer_head direct IO code Matthew Bobrowski
2019-09-16 12:06 ` Christoph Hellwig
2019-09-16 12:53 ` Matthew Bobrowski
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=20190917124859.GC17286@bobrowski \
--to=mbobrowski@mbobrowski.org \
--cc=adilger.kernel@dilger.ca \
--cc=darrick.wong@oracle.com \
--cc=david@fromorbit.com \
--cc=hch@infradead.org \
--cc=jack@suse.cz \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=tytso@mit.edu \
/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).