From: Mel Gorman <mgorman@suse.de>
To: Josef Bacik <jbacik@fb.com>
Cc: linux-fsdevel@vger.kernel.org, david@fromorbit.com,
viro@zeniv.linux.org.uk, jack@suse.cz,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH 0/9] Sync and VFS scalability improvements
Date: Wed, 25 Mar 2015 11:18:58 +0000 [thread overview]
Message-ID: <20150325111858.GA5599@suse.de> (raw)
In-Reply-To: <1426016724-23912-1-git-send-email-jbacik@fb.com>
On Tue, Mar 10, 2015 at 03:45:15PM -0400, Josef Bacik wrote:
> These are patches that Dave Chinner wrote two years ago that are still very much
> needed today. I recently ran into a problem where I had millions of inodes that
> needed to be evicted at unmount time and it soft locked up the box and kept any
> other file system from doing work. These patches fix this problem by breaking
> the inode_sb_list_lock into per-sb, and then converting the per sb inode list
> into a list_lru for better scalability.
>
I'm not reviewing this series, I just heard it mentioned at LSF/MM and again
on a DAX thread and am taking a quick glance out of curiousity. FWIW, I saw
the lockup-on-umount warning every time I ran Dave's xfsrepair workload when
unmounting after running fsmark. I didn't report it because Dave mentioned
at Vault that it was a known problem and someone was looking at it already.
--
Mel Gorman
SUSE Labs
prev parent reply other threads:[~2015-03-25 11:19 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-03-10 19:45 [PATCH 0/9] Sync and VFS scalability improvements Josef Bacik
2015-03-10 19:45 ` [PATCH 1/9] writeback: plug writeback at a high level Josef Bacik
2015-03-10 19:45 ` [PATCH 2/9] inode: add IOP_NOTHASHED to avoid inode hash lock in evict Josef Bacik
2015-03-12 9:52 ` Al Viro
2015-03-12 12:18 ` [PATCH] inode: add hlist_fake to avoid the " Josef Bacik
2015-03-12 12:20 ` [PATCH] inode: add hlist_fake to avoid the inode hash lock in evict V2 Josef Bacik
2015-03-14 7:00 ` Jan Kara
2015-03-12 12:24 ` [PATCH 2/9] inode: add IOP_NOTHASHED to avoid inode hash lock in evict Josef Bacik
2015-03-10 19:45 ` [PATCH 3/9] inode: convert inode_sb_list_lock to per-sb Josef Bacik
2015-03-10 19:45 ` [PATCH 4/9] sync: serialise per-superblock sync operations Josef Bacik
2015-03-10 19:45 ` [PATCH 5/9] inode: rename i_wb_list to i_io_list Josef Bacik
2015-03-10 19:45 ` [PATCH 6/9] bdi: add a new writeback list for sync Josef Bacik
2015-03-16 10:14 ` Jan Kara
2015-03-10 19:45 ` [PATCH 7/9] writeback: periodically trim the writeback list Josef Bacik
2015-03-16 10:16 ` Jan Kara
2015-03-16 11:43 ` Jan Kara
2015-03-10 19:45 ` [PATCH 8/9] inode: convert per-sb inode list to a list_lru Josef Bacik
2015-03-16 12:27 ` Jan Kara
2015-03-16 15:34 ` Josef Bacik
2015-03-16 15:48 ` Jan Kara
2015-03-10 19:45 ` [PATCH 9/9] inode: don't softlockup when evicting inodes Josef Bacik
2015-03-16 12:31 ` Jan Kara
2015-03-16 11:39 ` [PATCH 0/9] Sync and VFS scalability improvements Jan Kara
2015-03-25 11:18 ` Mel Gorman [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=20150325111858.GA5599@suse.de \
--to=mgorman@suse.de \
--cc=david@fromorbit.com \
--cc=jack@suse.cz \
--cc=jbacik@fb.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@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 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.