From: David Sterba <dsterba@suse.cz>
To: WenRuo Qu <wqu@suse.com>
Cc: "dsterba@suse.cz" <dsterba@suse.cz>,
"linux-btrfs@vger.kernel.org" <linux-btrfs@vger.kernel.org>,
James Harvey <jamespharvey20@gmail.com>
Subject: Re: [PATCH v2] btrfs: inode: Don't compress if NODATASUM or NODATACOW set
Date: Fri, 5 Jul 2019 18:38:28 +0200 [thread overview]
Message-ID: <20190705163828.GC20977@twin.jikos.cz> (raw)
In-Reply-To: <7eedec47-5ab1-0d39-7cb0-e7ab75b2424a@suse.com>
On Thu, Jul 04, 2019 at 11:51:38PM +0000, WenRuo Qu wrote:
> >> @@ -1630,7 +1655,8 @@ int btrfs_run_delalloc_range(struct inode *inode, struct page *locked_page,
> >> } else if (BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC && !force_cow) {
> >> ret = run_delalloc_nocow(inode, locked_page, start, end,
> >> page_started, 0, nr_written);
> >> - } else if (!inode_need_compress(inode, start, end)) {
> >> + } else if (!inode_can_compress(inode) ||
> >> + !inode_need_compress(inode, start, end)) {
> >
> > Well, that's not excatly what I expected, but because this is an
> > important fix I won't object now and add it to 5.3 queue.
> >
>
> I know what you expect, single inode_can_compress().
Yeah, because need_compress calls the heuristic, and then again inside
compress_file_range. I found a few problems in the compression decision
logic that will address that but for now that's how things have been so
the fix is not making it worse.
> But still, we want to avoid hitting the compression routine, thus here
> we do extra inode_need_compress() check other than exiting in
> compress_file_extent().
That's right, some of the compression decisions can be made out of
compress_file_range, like NOCOMPRESS. Even the heuristics can let it
skip compression completely but when this does not reach
compress_file_range the NOCOMPRESS flag has no chance to be set.
And there are more issues.
prev parent reply other threads:[~2019-07-05 16:37 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-01 5:12 [PATCH v2] btrfs: inode: Don't compress if NODATASUM or NODATACOW set Qu Wenruo
2019-07-04 16:04 ` David Sterba
2019-07-04 23:51 ` WenRuo Qu
2019-07-05 16:38 ` David Sterba [this message]
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=20190705163828.GC20977@twin.jikos.cz \
--to=dsterba@suse.cz \
--cc=jamespharvey20@gmail.com \
--cc=linux-btrfs@vger.kernel.org \
--cc=wqu@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).