<?xml version="1.0"?>
<feed xmlns="http://www.w3.org/2005/Atom" xml:lang="en">
	<id>https://xfs.org/index.php?action=history&amp;feed=atom&amp;title=XFS_status_update_for_2009</id>
	<title>XFS status update for 2009 - Revision history</title>
	<link rel="self" type="application/atom+xml" href="https://xfs.org/index.php?action=history&amp;feed=atom&amp;title=XFS_status_update_for_2009"/>
	<link rel="alternate" type="text/html" href="https://xfs.org/index.php?title=XFS_status_update_for_2009&amp;action=history"/>
	<updated>2026-04-20T09:01:14Z</updated>
	<subtitle>Revision history for this page on the wiki</subtitle>
	<generator>MediaWiki 1.42.3</generator>
	<entry>
		<id>https://xfs.org/index.php?title=XFS_status_update_for_2009&amp;diff=2439&amp;oldid=prev</id>
		<title>Pibroch: New page: == XFS status update for December 2009 ==  December finally saw the long awaited release of Linux 2.6.32, which for XFS is mostly a bug fix release, with the major changes being various im...</title>
		<link rel="alternate" type="text/html" href="https://xfs.org/index.php?title=XFS_status_update_for_2009&amp;diff=2439&amp;oldid=prev"/>
		<updated>2012-03-19T06:14:14Z</updated>

		<summary type="html">&lt;p&gt;New page: == XFS status update for December 2009 ==  December finally saw the long awaited release of Linux 2.6.32, which for XFS is mostly a bug fix release, with the major changes being various im...&lt;/p&gt;
&lt;p&gt;&lt;b&gt;New page&lt;/b&gt;&lt;/p&gt;&lt;div&gt;== XFS status update for December 2009 ==&lt;br /&gt;
&lt;br /&gt;
December finally saw the long awaited release of Linux 2.6.32, which for&lt;br /&gt;
XFS is mostly a bug fix release, with the major changes being various&lt;br /&gt;
improvement to the sync path, including working around the expectation&lt;br /&gt;
from the grub boot loader where metadata is supposed to be after a sync()&lt;br /&gt;
system call.  Together with a refactoring of the inode allocator this&lt;br /&gt;
gives a nice diffstat for this kernel release:&lt;br /&gt;
&lt;br /&gt;
 46 files changed, 767 insertions(+), 1048 deletions(-)&lt;br /&gt;
&lt;br /&gt;
In the meantime development for the 2.6.33 has been going strong.  The&lt;br /&gt;
new event tracing code that allows to observe the inner workings of XFS&lt;br /&gt;
in production systems has finally been merged, with another patch to&lt;br /&gt;
reduce the size of the tracing code by using new upstream kernel features&lt;br /&gt;
posted for review.  Also a large patch series has been posted which&lt;br /&gt;
changes per-AG data to be looked up by a radix tree instead of the&lt;br /&gt;
existing array.  This works around possible deadlocks and user after&lt;br /&gt;
free issues during growfs, and prepares for removing a global (shared)&lt;br /&gt;
lock from the free space allocators.  In addition to that a wide range&lt;br /&gt;
of fixes has been posted and applied.&lt;br /&gt;
&lt;br /&gt;
Work on the userspace packages has been just as busy.  In mkfs.xfs the&lt;br /&gt;
lazy superblock counter feature has now been enabled by default for the&lt;br /&gt;
upcoming xfsprogs 3.1.0 release, which will require kernel 2.6.22 for&lt;br /&gt;
the default mkfs invocation.  Also for mkfs.xfs as patch was posted&lt;br /&gt;
to correct the automatic detection of 4 kilobyte sector drivers which&lt;br /&gt;
are expected to show up in large quantities the real work soon.  The&lt;br /&gt;
norepair mode in xfs_repair has been enhanced with additional freespace&lt;br /&gt;
btree correction checks from xfs_db and is now identical to xfs_check in&lt;br /&gt;
filesystem consistency checking coverage.  A temporary file permission&lt;br /&gt;
problems has been fixed in xfs_fsr, and the libhandle library has been&lt;br /&gt;
fixed to better deal with symbolic links.  In xfs_io a few commands&lt;br /&gt;
that were added years ago have finally been wired up to actually be&lt;br /&gt;
usable.  And last but not least xfsdump saw a fix to the time stamp&lt;br /&gt;
handling in the backup format and some usability and documentation&lt;br /&gt;
improvements to xfsinvutil.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for November 2009 ==&lt;br /&gt;
&lt;br /&gt;
November was a relatively slow month for XFS development.  The XFS tree&lt;br /&gt;
that is destined for the Linux 2.6.33 merge window saw a few fixes and&lt;br /&gt;
cleanups applied to it, and few important fixes still made it into the&lt;br /&gt;
last Linux 2.6.32 release candidates.  A few more patches including a&lt;br /&gt;
final version of the event tracing support for XFS were posted but not&lt;br /&gt;
reviewed yet.&lt;br /&gt;
&lt;br /&gt;
On the userspace side there has been a fair amount of xfsprogs activity.&lt;br /&gt;
The repair speedup patches have finally been merged into the main development&lt;br /&gt;
branch and a couple of other fixes to the various utilities made it in, too.&lt;br /&gt;
The xfstests test suite saw another new regression test suite and a build&lt;br /&gt;
system fix up.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for October 2009 ==&lt;br /&gt;
In October we saw the Linux 2.6.32 merge window with a major XFS update.&lt;br /&gt;
This update includes a refactoring of the inode allocator which also&lt;br /&gt;
allows for speedups for very large filesystems, major sync fixes, updates&lt;br /&gt;
to the fsync and O_SYNC handling which merge the two code paths into a single&lt;br /&gt;
and more efficient one, a workaround for the VFS time stamp behavior,&lt;br /&gt;
and of course various smaller fixes.  A couple of additional fixes have been&lt;br /&gt;
queued up for the next merge window.&lt;br /&gt;
&lt;br /&gt;
On the userspace side there has been a healthy activity on xfsprogs:  mkfs can&lt;br /&gt;
now discard unused sectors on SSDs and thinly provisioned storage devices and&lt;br /&gt;
use the more generic libblkid for topology information and filesystems detection&lt;br /&gt;
instead of the older libdisk, and the build system gained some updates to&lt;br /&gt;
make the source package generation simpler and shared for different package&lt;br /&gt;
types.  A patch has been out to the list but yet committed to add symbol&lt;br /&gt;
versioning to the libhandle library to make future ABI additions easier.&lt;br /&gt;
The xfstests package only saw some minor activity with a new test case&lt;br /&gt;
and small build system fixes.&lt;br /&gt;
&lt;br /&gt;
New minor releases of xfsprogs and xfsdump were tagged but not formally&lt;br /&gt;
released after additional discussion.  Instead a new major xfsprogs release&lt;br /&gt;
is planned for next month.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for September 2009 ==&lt;br /&gt;
&lt;br /&gt;
In September the Linux 2.6.31 kernel was finally released, including another&lt;br /&gt;
last minute XFS fix for the swapext (defragmentation) compat ioctl handler.&lt;br /&gt;
The final patch from 2.6.30 to 2.6.31 shows the following impressive diffstat&lt;br /&gt;
for XFS:&lt;br /&gt;
&lt;br /&gt;
   55 files changed, 1476 insertions(+), 2269 deletions(-)&lt;br /&gt;
&lt;br /&gt;
The 2.6.32 merge window started with a large XFS merge that included changes&lt;br /&gt;
to the inode allocator, and a few smaller fixes.  New versions of the sync&lt;br /&gt;
and time stamp fixes as well as the event tracing support have been posted&lt;br /&gt;
in September but not yet merged into the XFS development tree and/or mainline.&lt;br /&gt;
&lt;br /&gt;
On the userspace side a large patch series to reduce the memory usage in&lt;br /&gt;
xfs_repair to acceptable levels was posted, but not yet merged.  A new xfs_df&lt;br /&gt;
shell script to measure use of the on disk space was posted but not yet&lt;br /&gt;
merged pending some minor review comments and a missing man page.  In addition&lt;br /&gt;
we saw the usual amount of smaller fixes and cleanups.&lt;br /&gt;
&lt;br /&gt;
Also this month Felix Blyakher resigned from his post as XFS maintainer and handed off to Alex Elder.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for August 2009 ==&lt;br /&gt;
&lt;br /&gt;
In August the Linux 2.6.31 kernel has still been in the release candidate&lt;br /&gt;
stage, but a couple of important XFS fixes made it in time for the release,&lt;br /&gt;
including a fix for the inode cache races with NFS workloads that have&lt;br /&gt;
plagued us for a long time.&lt;br /&gt;
&lt;br /&gt;
The list saw various patches destined for the Linux 2.6.32 merge window,&lt;br /&gt;
including a merge of the fsync and O_SYNC handling code to address various&lt;br /&gt;
issues with the latter, a workaround for deficits in the timestamp handling&lt;br /&gt;
interface between the VFS and filesystems, a repost of the sync improvements&lt;br /&gt;
patch series and various smaller patches.&lt;br /&gt;
&lt;br /&gt;
August also saw the minor 3.0.3 release of xfsprogs which collects smaller&lt;br /&gt;
fixes to the various tools and most importantly a fix to allow xfsprogs to&lt;br /&gt;
work again on SPARC and other strict alignment handling which regressed a&lt;br /&gt;
few releases ago.  The xfstests repository saw a few new test cases and a&lt;br /&gt;
various small improvements.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for July 2009 ==&lt;br /&gt;
&lt;br /&gt;
As a traditional summer vacation month July has not seen a lot of XFS&lt;br /&gt;
activity.  The mainline 2.6.31 kernel made it to the 5th release candidate&lt;br /&gt;
but besides a few kernel-wide patches touching XFS the only activity were&lt;br /&gt;
two small patches fixing a bug in FIEMAP and working around writeback&lt;br /&gt;
performance problems in the VM.&lt;br /&gt;
&lt;br /&gt;
A few more patches were posted to the list but haven&amp;#039;t been merged yet.&lt;br /&gt;
Two big patch series deal with theoretically possible deadlocks due to&lt;br /&gt;
locks taken in reclaim contexts, which are now detected by lockdep.&lt;br /&gt;
&lt;br /&gt;
The pace on the userspace side has been slow.  There have been a couple&lt;br /&gt;
of fixes to xfs_repair and xfs_db, and xfstests grew a few more testcases.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for June 2009 ==&lt;br /&gt;
&lt;br /&gt;
On June 9th we finally saw the release of Linux 2.6.30.  For XFS&lt;br /&gt;
this release mostly contains the improved ENOSPC handling, but also&lt;br /&gt;
various smaller bugfixes and lots of cleanups.  The code size of XFS&lt;br /&gt;
decreased again by 500 lines of code in this release.&lt;br /&gt;
&lt;br /&gt;
The Linux 2.6.31 merge opened in the mid of the month and some big XFS&lt;br /&gt;
changes have been pushed: A removal of the quotaops&lt;br /&gt;
infrastructure which simplifies the quota implementation, the switch&lt;br /&gt;
from XFS&amp;#039;s own Posix ACL implementation to the generic one shared&lt;br /&gt;
by various other filesystems which also supports in-memory caching of&lt;br /&gt;
ACLs and another incremental refactoring of the sync code.&lt;br /&gt;
&lt;br /&gt;
A patch to better track dirty inodes and work around issues in the&lt;br /&gt;
way the VFS updates the access time stamp on inodes has been reposted&lt;br /&gt;
and discussed. Another patch to converting the existing XFS tracing&lt;br /&gt;
infrastructure to use the ftrace even tracer has been posted.&lt;br /&gt;
&lt;br /&gt;
On the userspace side there have been a few updates to xfsprogs, including&lt;br /&gt;
some repair fixes and a new fallocate command for xfs_io.  There were&lt;br /&gt;
major updates for xfstests:  The existing aio-dio-regress testsuite has&lt;br /&gt;
been merged into xfstests, and various changes went into the tree to make&lt;br /&gt;
xfstests better suitable for use with other filesystems.&lt;br /&gt;
&lt;br /&gt;
The attr and acl projects which have been traditionally been hosted&lt;br /&gt;
as part of the XFS userspace utilities have now been split into a separate&lt;br /&gt;
project maintained by Andreas Gruenbacher, who has been doing most of&lt;br /&gt;
the work on it, and moved to the Savannah hosting platform.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for May 2009 ==&lt;br /&gt;
&lt;br /&gt;
In May Linux 2.6.30 was getting close to be released, and various&lt;br /&gt;
important XFS fixes made it during the latest release candidates.&lt;br /&gt;
In the meantime some big patch series to rework the sync code and&lt;br /&gt;
the inode allocator have been posted for the next merge window.&lt;br /&gt;
&lt;br /&gt;
On the userspace side xfsprogs and xfsdump 3.0.1 were finally released,&lt;br /&gt;
quickly followed by 3.0.2 releases with updated Debian packaging.&lt;br /&gt;
After that various small patches that were held back made it into xfsprogs.&lt;br /&gt;
A patch to add the xfs_reno tool which allows to move inodes around to&lt;br /&gt;
fit into 32 bit inode number space has been posted which is also one&lt;br /&gt;
central aspect of future online shrinking support.&lt;br /&gt;
&lt;br /&gt;
There has been major activity on xfstests including adding generic&lt;br /&gt;
filesystems support to allow running tests that aren&amp;#039;t XFS-specific on&lt;br /&gt;
any Linux filesystems.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for April 2009 ==&lt;br /&gt;
&lt;br /&gt;
In April development for Linux 2.6.30 was in full motion.  A patchset to correct flushing of delayed allocations with near full filesystems has been committed in early April, as well as various smaller fixes. A patch series to improve the behavior of sys_sync has been posted but is waiting for VFS changes queued for Linux 2.6.31.&lt;br /&gt;
&lt;br /&gt;
On the userspace side xfsprogs and xfsdump 3.0.1 have managed to split their release dates into May again after a lot of last-minute build system updates.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for March 2009 ==&lt;br /&gt;
&lt;br /&gt;
Linux 2.6.29 has been released which includes major XFS updates like the&lt;br /&gt;
new generic btree code, a fully functional 32bit compat ioctl implementation&lt;br /&gt;
and the new combined XFS and Linux inode.  (See previous status reports&lt;br /&gt;
for more details). A patch series to improve correctness and performance&lt;br /&gt;
has been posted but not yet applied.  Various minor fixes and cleanups&lt;br /&gt;
have been sent to Linus for 2.6.30 which looks like it will be a minor&lt;br /&gt;
release for XFS after the big churn in 2.6.29.&lt;br /&gt;
&lt;br /&gt;
On userspace a lot of time has been spent on fixing and improving the&lt;br /&gt;
build system shared by the various XFS utilities as well as various smaller&lt;br /&gt;
improvements leading to the xfsprogs and xfsdump 3.0.1 releases which are&lt;br /&gt;
still outstanding.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for February 2009 ==&lt;br /&gt;
&lt;br /&gt;
In February various smaller fixes have been sent to Linus for 2.6.29,&lt;br /&gt;
including a revert of the faster vmap APIs which don&amp;#039;t seem to be quite&lt;br /&gt;
ready yet on the VM side.  At the same time various patches have been&lt;br /&gt;
queued up for 2.6.30, with another big batch pending.  There also has&lt;br /&gt;
been a repost of the CRC patch series, including support for a new,&lt;br /&gt;
larger inode core.&lt;br /&gt;
&lt;br /&gt;
SGI released various bits of work in progress from former employees&lt;br /&gt;
that will be extremely helpful for the future development of XFS,&lt;br /&gt;
thanks a lot to Mark Goodwin for making this happen.&lt;br /&gt;
&lt;br /&gt;
On the userspace side the long awaited 3.0.0 releases of xfsprogs and&lt;br /&gt;
xfsdump finally happened early in the month, accompanied by a 2.2.9&lt;br /&gt;
release of the dmapi userspace.  There have been some issues with packaging&lt;br /&gt;
so a new minor release might follow soon.&lt;br /&gt;
&lt;br /&gt;
The xfs_irecover tool has been relicensed so that it can be merged into&lt;br /&gt;
the GPLv2 codebase of xfsprogs, but the actual integration work hasn&amp;#039;t&lt;br /&gt;
happened yet.&lt;br /&gt;
&lt;br /&gt;
Important bits of XFS documentation that have been available on the XFS&lt;br /&gt;
website in PDF form have been released in the document source form under&lt;br /&gt;
the Creative Commons license so that they can be updated as a community&lt;br /&gt;
effort, and checked into a public git tree.&lt;br /&gt;
&lt;br /&gt;
== XFS status update for January 2009 ==&lt;br /&gt;
&lt;br /&gt;
January has been an extremely busy month on the userspace front.  Many&lt;br /&gt;
smaller and medium updates went into xfsprogs, xfstests and to a lesser&lt;br /&gt;
extent xfsdump.  xfsprogs and xfsdump are ramping up for getting a 3.0.0&lt;br /&gt;
release out in early February which will include the first major re-sync&lt;br /&gt;
with the kernel code in libxfs, a cleanup of the exported library interfaces&lt;br /&gt;
and the move of two tools (xfs_fsr and xfs_estimate) from the xfsdump&lt;br /&gt;
package to xfsprogs.  After this the xfsprogs package will contain all&lt;br /&gt;
tools that use internal libxfs interfaces which fortunately equates to those&lt;br /&gt;
needed for normal administration.  The xfsdump package now only contains&lt;br /&gt;
the xfsdump/xfsrestore tools needed for backing up and restoring XFS&lt;br /&gt;
filesystems.  In addition it grew a fix to support dump/restore on systems&lt;br /&gt;
with a 64k page size.  A large number of acl/attr package patches was&lt;br /&gt;
posted to the list, but pending a possible split of these packages from the&lt;br /&gt;
XFS project these weren&amp;#039;t processed yet.&lt;br /&gt;
&lt;br /&gt;
On the kernel side the big excitement in January was an in-memory corruption&lt;br /&gt;
introduced in the btree refactoring which hit people running 32bit platforms&lt;br /&gt;
without support for large block devices.  This issue was fixed and pushed&lt;br /&gt;
to the 2.6.29 development tree after a long collaborative debugging effort&lt;br /&gt;
at linux.conf.au.  Besides that about a dozen minor fixes were pushed to&lt;br /&gt;
2.6.29 and the first batch of misc patches for the 2.6.30 release cycle&lt;br /&gt;
was sent out.&lt;br /&gt;
&lt;br /&gt;
At the end of December the SGI group in Melbourne which the previous&lt;br /&gt;
XFS maintainer and some other developers worked for has been closed down&lt;br /&gt;
and they will be missed greatly.  As a result maintainership has been passed&lt;br /&gt;
on in a way that has been slightly controversial in the community, and the&lt;br /&gt;
first patchset of work in progress in Melbourne have been posted to the list&lt;br /&gt;
to be picked up by others.&lt;br /&gt;
&lt;br /&gt;
The xfs.org wiki has gotten a little facelift on it&amp;#039;s front page making it&lt;br /&gt;
a lot easier to read.&lt;/div&gt;</summary>
		<author><name>Pibroch</name></author>
	</entry>
</feed>