Hello all,
This is the first weekly CVS release in 2009.
As I began developing AUFS2 (cf. the URL), this is also the final
regular weekly CVS release.
http://sourceforge.net/mailarchive/forum.php?thread_name=7408.1232332832%40jroblforum_name=aufs-users
When a bug is reported or a new feature is
Hi all,
This is the final regular release in this year. If something wrong
happens, I may release a fixed version as an emergency release.
Have nice holidays!
J. R. Okajima
o bugfix
- bugfix(completed): stop recursive si lock, reported by Louis Rilling,
refine the readdir call from NFSD
o news
- introduce two new mount options sum/nosum for statfs(2). sum returns
the total number of blocks/inodes of all branches.
Thanks again for this one. Will try it as soon as I get
the chance.
Leslie
-
This
o bugfix
- bugfix: the dir inode may remain without the corresponding dentry,
regression on 20080929, reported by James.
- bugfix: initilize i_op for special inodes, regression on 20080804,
reported by Kris Warkentin.
- bugfix: use the passed mount option flags insted of the currently set
o bugfix
- bugfix: NFS exporting and udba=inotify caused deadlock, make knfsd not
to wait for the completion of all queued tasks, reported by Louis
Rilling.
+ remove AuLock_FLUSH flag.
+ stop waiting in case of knfsd.
- bugfix: deadlock for changing the attributes of the root dir on
o news
- begin supporting linux-2.6.27.
o bugfix
- bugfix: race condition in rename(), fix the locking order of
s_vfs_rename_mutex, reported by Louis Rilling.
+ introduce a new flag VFS_RENAME for au_pin.
+ new abstract function vfsub_is_rename_mutex_locked().
+ test the lock for
o bugfix
- bugfix: restore lockdef_off/on(), a regression on 20080922.
- bugfix: check if the destroying inode is dead when maintaining the
inode generation.
- bugfix: when the inode/dentry was still cached, check its generation.
o news
- support mnt_want/drop_write() in linux-2.6.26 and
o bugfix
- bugfix (fs/aufs, fs/aufs25): introduce a new mutex to prevent a new
inode number for hardlinks under different parnet dir from a race
condition.
+ (fs/aufs25) a new structure au_xino_file.
- bugfix (fs/aufs, fs/aufs25): support an error case of opening a dir.
- bugfix (fs/aufs,
o news
- new configuration CONFIG_AUFS_UNIONFS2[23]_PATCH to support
unionfs-patched kernel, since several users think it is a problem of
aufs (of course it is not actually).
+ new abstract macros for splice functions.
+ new abstract function vfsub_copy_inode_size(), move it from cpup.c
o news
- begin supporting 2.6.26-rc2-mm1.
+ -umount_begin() will not be called since a condition is added in
VFS. aufs used this operation to maintain the internal vfsmount
pointer and it becomes impossible now.
* make some features ('nowait' workqueue job and xino truncation)
o bugfix
- (fs/aufs, fs/aufs25) bugfix: assign the return value of realloc and
update the pointer, reported and patched by Mark A. Grondona.
- (fs/aufs, fs/aufs25) bugfix: a race condition in page fault, reported
by Mark A. Grondona.
+ an explicit lock for lower -fault() (or -nopage()).
+
- rename si_mnt to si_mntcache, and a new accessor au_mntcache_get()
(unfinished).
+ remove au_update_mnt().
+ unset si_mntcache in aufs_umount_begin().
+ replace struct super_block in struck au_wkinfo by struct vfsmount.
- new internal configuration CONFIG_AUFS_LOCAL which
o bugfix
- (fs/aufs, fs/aufs25) bugfix: ignore an event for a being removed inode.
o news
- introduce a kobject to manage the lifetime of sbinfo, based upon a
patch from Jeff Mahoney.
+ remove CONFIG_AUFS_SYSAUFS from fs/aufs25.
+ compile sysaufs.c unconditionally.
+ new file sysfs.c
Jeff Mahoney:
Well, sysfs is an easy path once you've embraced the kobject/kset model.
:::
You have succeeded push me into the sea of kobject/sysfs.
I have almost completed rewriting the source files which are,
- the lifetime of sbinfo is handled by kobject unconditinally, which
means
I'm not sure if I understand you right, but I believe you are wondering if
CONFIG_SYSFS is usually enabled in Linux distributions.
I think it is enabled on ALL.
Normally, it will not allow you to disable sysfs. When user manually edits
.config file to use CONFIG_SYSFS=n, it doesn't let him -
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
[EMAIL PROTECTED] wrote:
o news
- begin supporting linux-2.6.25 by ./fs/aufs25. now ./fs/aufs supports
linux-2.6.16 - 2.6.24.
:::
Sorry, forgot to write one thing.
Supporting linux-2.6.25 is triggered by Jeff Mahoney and based upon
Hello Tomas and Jeff,
Tomas M:
I'm not sure if I understand you right, but I believe you are wondering if
CONFIG_SYSFS is usually enabled in Linux distributions.
I think it is enabled on ALL.
Agreed.
The only way to disable sysfs is when you set CONFIG_EMBEDDED=y, which is
IMHO very
The smaller code size applies to both cases, just not entirely as
small due to the size of the kobject. It's a pretty small structure,
of which we use one per superblock. The scale makes any savings
negligible.
The sysfs code is all static inline calls, so it carries no footprint
when
Jeff Mahoney:
The smaller code size applies to both cases, just not entirely as
small due to the size of the kobject. It's a pretty small structure,
of which we use one per superblock. The scale makes any savings
negligible.
The sysfs code is all static inline calls, so it carries no
o news
- begin supporting linux-2.6.25 by ./fs/aufs25. now ./fs/aufs supports
linux-2.6.16 - 2.6.24.
+ pass the real value of vfsmount* for dentry_open().
+ remove iget_locked() and -read_inode(), and replace them by a new
function au_iget_locked().
+ drop IN_DELETE_SELF and
o news
- begin supporting linux-2.6.25 by ./fs/aufs25. now ./fs/aufs supports
linux-2.6.16 - 2.6.24.
:::
Sorry, forgot to write one thing.
Supporting linux-2.6.25 is triggered by Jeff Mahoney and based upon
patches from him which were helpful.
Still CONFIG_AUFS_SYSAUFS remains
o bugfix
- bugfix: pass DIR flag only to di_write_lock2_child().
o misc
- remove dlgt.h and robr.h.
- revert 'struct au_opts' with bit-fields to 'unsigned int' with
bit-shift.
+ extract a part of aufs_show_options() and make a new function
au_show_wbr_create().
+ new enum
o news
- add some articles from users in personal mail.
If you don't want to list your name here, write to me.
o bugfix
- bugfix: update inode-i_ctime after unlink().
- bugfix: update inode-i_ctime after rename().
o misc
- consolidate '#ifdef's.
+ automatic configuration for internal use,
o news
- unveil CONFIG_AUFS_SHWH.
- re-write sysfs interface based upon a patch by Jeff
Mahoney. (testing)
+ todo: support files larger than PAGE_SIZE.
+ remove CONFIG_AUFS_SYSFS_GET_DENTRY_PATCH.
+ remove sysfs_get_dentry.patch
+ move au_sbilist_mtx and au_sbilist from sysaufs.c to
o bugfix
- bugfix: udba=inotify, reported by Klaus Fuerstberger (still testing).
+ in case of copy-up, clear the xino for lower copyup-src. because
it prevents aufs from refreshing the inode number which is
udba-ed.
+ flush postproc.
+ skip readdir for udba-removed dir.
+ stop
o news
- begin re-writing sysfs interface based upon a patch by Jeff
Mahoney. (still working)
When this rework is finished, /sys/fs/aufs/brs will be gone.
Instead, /sys/fs/aufs/mnt_/brN will come.
Junjiro Okajima
o bugfix
- bugfix: in case of testing super-io, stop forcing permission-check
for nfs branch, reported by Bruno Cesar Ribas.
+ extract a part of forcing permission check from au_test_perm(),
and create a new function au_test_perm_sio().
- bugfix: support a previous failure of setting
o bugfix
- bugfix: support complex whiteout/diropq case, when an existed upper
entry is not a directory, skip the test for diropq, reported by
Michael Towers.
- bugfix: a race condtion between UDBA and aufs.
+ check inode generation too
+ update file generation too
+ merge updating
o bugfix
- bugfix: revert the changes about mmap handling in 20080211, but
handles fuse individually.
+ remove reiser4.patch.
- bugfix: support ftruncate(2) (completed).
o news
- support fallocate(), actually aufs doesn't implement it as many
other filesystems do, reported by Yair Yarom.
-
o bugfix
- bugfix: reset f_mapping for memory mapped files.
+ new function au_vm_ops().
+ reset f_mapping in mmap().
+ use f_mapping-host instead of f_dentry-d_inode in sync().
o news
- enable AUFS_SUPER_MAGIC.
o misc
- test discarding cached fuse info.
- use f_mapping-host instead of
o bugfix
- bugfix: support for lazy unmount on linux-2.6.18 and later, reported
by Jorgen P. Tjerno.
- bugfix: a race condition between udba and aufs (unfnished).
o misc
- debugging a race condition between udba and aufs.
Junjiro Okajima
o news
- new sample chroot-ed/jail environment useing aufs.
o bugfix
- bugfix: consolidate make targets for parallel build (make -j3),
reported and tested by Tommy[D] and mike.
- bugfix: ignore an inotify event for the grand parent dir timestamps.
- bugfix: stop a race condition between
On Thu January 10 2008 00:23, [EMAIL PROTECTED] wrote:
Michael S. Zick:
Found it!
Congratulation.
Which course would you like to see - a simple rename everywhere,
or major change to look it up in the header file rather than grep for it?
There should be a CONFIG_VSERVER (or such
Hello Michael,
Michael S. Zick:
There is no #define TEMPFS_MAGIC in mm/shmem.c of 2.6.24-rc6
The comment about it being used by glibc remains, but the define is _gone_.
I downloaded
ftp.kernel.org/pub/linux/kernel/v2.6/testing/linux-2.6.24-rc6.tar.bz2
just now, and found there still exists
On Wed January 9 2008 15:35, [EMAIL PROTECTED] wrote:
Hello Michael,
Michael S. Zick:
There is no #define TEMPFS_MAGIC in mm/shmem.c of 2.6.24-rc6
The comment about it being used by glibc remains, but the define is _gone_.
I downloaded
Found it!
TMPFS_SUPER_MAGIC define added to: include/linux/shmem_fs.h
And the changes earlier posted made:
By the pre-release Linux-VServer patch (either of these):
patch-2.6.24-rc6-vs2.2.0.5.0.4.diff
patch-2.6.24-rc6-vs2.2.0.5.0.6-pre.diff
http://vserver.13thfloor.at/Experimental/ - bottom of
o bugfix
- bugfix: AUFS_LHASH_PATCH depends on AUFS_FAKE_DM = n, reported by
Russell Harmon and Heinrich Rebehn.
- bugfix: a xino-truncation thread must not wait for other threads.
o misc
- add a note about setting 'n.'
- move all the declaration from super.h to branch.h.
- rename struct xino
Hello.
May I ask you, what is the trunc_xino useful for?
I didn't find anything in aufs.5 documentation,
only trunc_xib and notrunc_xib, but that is perhaps
something else, am I right?
Thank you
Tomas M
[EMAIL PROTECTED] wrote:
o bugfix
- bugfix: a race in revalidating a dentry, reported
Hi,
Tomas M:
May I ask you, what is the trunc_xino useful for?
I didn't find anything in aufs.5 documentation,
only trunc_xib and notrunc_xib, but that is perhaps
something else, am I right?
Right.
Because I don't know whether this feature is required or not, I didn't
write the
Hello,
tried Monday release of aufs with 2.6.24-rc3 and doesn't compile for me:
sysaufs.c: error: implicit declaration of subsys_put and subsys_get
(in function reg and unreg)
Should I wait for 2.6.24 final?
Thank you
Tomas M
[EMAIL PROTECTED] wrote:
o news
- introduce a new semantics for
o news
- introduce a new semantics for link(2) and rename(2), and three
policies to select one among multiple writable branches (completed).
+ implement a new create policy 'mfsrr.'
(quote from the aufs manual)
Policies to Select One among Multiple Writable Branches
Aufs has some
o security
- security enhancement or possible bug fix: support FMODE_EXEC and
{deny,allow}_write_access() for an executable and put_write_access()
for an open file for writing on a branch. for linux-2.6.17 and later.
o bugfix
- bugfix: replace a direct access to dentry-d_parent by
o bugfix
- bugfix: at decoding a nfs handle, in the slowest way which will be
invoked after other two method's failure, aufs generates the path
for the decode target inode. and there was a bug in generating the
path. when the inode is decoded by the two preceeding methods, you
will not
o bugfix
- bugfix: name collision on ia64, reported and patched by Julian
Andres Klode, Daniel Baumann, Otavio Salvador and dann frazier (I
replaced si_flags by au_si_flags insead of aufs_si_flags).
- bugfix: when aufs is liked statically, /tmp/aufsvars.mk is
unnecessary, reported by Tapani
o bugfix
- bugfix: whiteout list may be obsoleted while aufs was sleeping,
reported by Torsten Luettgert.
o news
- begin supporting linux-2.6.23, still testing on rc6.
+ support splice(2).
+ introduce aufs_fault().
+ new patches and configurations,
* splice-2.6.23.patch +
o news
- the dlgt option is revealed.
o bugfix
- bugfix: handling whiteout in dlgt mode.
- bugfix: support umask in dlgt mode.
Junjiro Okajima
--
Index: fs/aufs/i_op_add.c
Index: fs/aufs/i_op_ren.c
- bugfix: handling
o current problems
- setting inode number which was previously assigned (broken ino
msg) (fixed).
- getdents(2) returns nothing. (fixed)
- cpup_wh_file() failure. (fixed)
o bugfix
- bugfix: cpup whiteout which was called #2 in last ci.
o news
- begin supporting linux-2.6.22
+ introduce
o current problems
- setting inode number which was previously assigned (broken ino
msg) (testing).
- getdents(2) returns nothing. this may involve two problems. (it was
one. testing)
- cpup_wh_file() failure. (testing)
o bugfix
- bugfix: force rewind at re-initializing vdir, reported by
o current problem
- isolated inode survived deleting a branch (still testing).
- getdents(2) returns nothing.
- cpup_wh_file() failure.
- setting inode number which was previously assigned (broken ino
msg).
- hang at reboot/shutdown (it may not be an aufs problem).
o bugfix
- bugfix: skip
o bugfix
- bugfix: declare 'signed char', a portability problem reported by
Bertrand D.
- bugfix: strict branch index check at the first adding, reported by
Bertrand D.
- bugfix: unlink a whiteout under a dir who has a sticky bit, reported
by Igor Karasynskyi.
o misc
- warn about
o news
- the unionctl script is moved from util/ to sample/, and will not be
supported any more.
o bugfix
- bugfix: support the moved root branch.
o misc
- testing is_subdir() kernel internal function.
Junjiro Okajima
--
Hi Junjiro,
Am Montag, 28. Mai 2007 04:04 schrieb [EMAIL PROTECTED]:
o misc
- testing linux-vserver, new abstract function do_vfsub_unlink().
great!
FYI, aufs works without problems on linux-vserver with the patches I sent to
the list some times ago. And it would be very good to see a clean
Hi Wilhelm,
Wilhelm Meier:
FYI, aufs works without problems on linux-vserver with the patches I sent to
the list some times ago. And it would be very good to see a clean solution
for linux-vserver incorporated into aufs.
I can guess supporting the linux-vserver interface (or function
Am Freitag, 1. Juni 2007 09:01 schrieb [EMAIL PROTECTED]:
Hi Wilhelm,
Wilhelm Meier:
FYI, aufs works without problems on linux-vserver with the patches I sent
to the list some times ago. And it would be very good to see a clean
solution for linux-vserver incorporated into aufs.
I can
o pre-announce
As I had announced a few months ago, the unionctl script will not be
supported.
The script the description about it in the aufs manual will be moved
from aufs_cvs/aufs/util to aufs_cvs/aufs/sample.
And the symlink under aufs_cvs/aufs/ will not be created by removing
it from
o misc
- flush all the scheduled/nowait tasks at umouning and remounting.
- refine the nowait task queuing.
- support the enqueue error in workqueue.
Junjiro Okajima
--
Index: fs/aufs/cpup.c
Index: fs/aufs/dentry.c
Index:
Sandino Flores Moreno:
... on linux 2.6.21.1, I got this error:
Sorry, I may be confused 2.6.21 with 2.6.22.
Please try this patch. If it succeeds, I will update CVS.
Junjiro Okajima
--
Index: fs/aufs/opts.h
o bugfix
- bugfix: file revalidation after rename(tgt).
- strict inode lock in fsync() before 2.6.17.
o news
- restore rdacahe option, which was dropped by mistake.
- print the depth of generic workqueue in sysfs/aufs/stat.
- testing initramfs.
- still testing udba=inotify.
- simplified
o bugfix
- bugfix: introduce dcsub.[ch] to handle dentry cache.
- bugfix: use readlink(1) to get cwd, in auplink, mount.aufs and
unionctl, reported by Torsten Luettgert. Actually, to use readlink
was suggested by Tomas M last year.
o news
- support initramfs.
- verbose warning message for
Marty Rosenberg:
To export aufs, you need to specify fsid option in your exports. Please
read the manual.
Sorry, this seems rather trivial, but I'm still unsure of how to get the
fsid of a given file system.
You can set it whatever you like. But it must be unique number.
more or less
o news
- introduce new branch attribute nolwh.
- refine parsing options at mount/remount time, reported by Sandino
Flores Moreno.
+ remove whiteout-base and whplink-dir in nopllink and br_mod(),
but br_del().
+ increment sbinfo generation in a new function
o bugfix
- bugfix: forgot dput in a error case.
o news
- introduce wkq_nowait().
- refine udba=inotify, still testing.
- extract some functions from sbinfo.c, create a new file plink.c.
Junjiro Okajima
--
Index:
Hello,
I think the Tuesday's release can't be compiled.
At least it doesn't work for me.
Using make -f local.mk, no other changes. Kernel 2.6.20:
start cut ---
make -C /lib/modules/2.6.20/build M=/usr/src/a/fs/aufs modules
make[1]: Entering directory `/usr/src/linux-2.6.20'
WARNING:
Tomas M:
I think the Tuesday's release can't be compiled.
:::
/usr/src/a/fs/aufs/misc.h:150: error: dereferencing pointer to
incomplete type
Sorry.
It was due to my debug environment.
Please update your CVS tree and try again.
I have just updated misc.h.
Junjiro Okajima
Marty Rosenberg:
I'd like to apologize for not giving enough information in the past, I'll try
to be a bit more verbose this time.
Config:
I've patched a vanilla 2.6.18.8 kernel with this tuesday's release.
M Another unionfs
--- These options are generated automatically for 2.6.18.8
65 matches
Mail list logo