The following patch stops NFS sillyname renames and umounts from racing. I have a test script does the following: 1) start nfs server 2) mount loopback 3) open file in background 4) remove file 5) stop nfs server 6) kill -9 process which has file open 7) restart nfs server 8) umount looback mount. After umount I got the "VFS: Busy inodes after unmount" message because the processing of the rename has not finished. Below is a patch that the ...

Signed-off-by: Erez Zadok <ezk@cs.sunysb.edu> Acked-by: Herton Ronaldo Krzesinski <herton@mandriva.com> --- fs/unionfs/union.h | 39 +++++++++++++++++++++++++++++++++------ 1 files changed, 33 insertions(+), 6 deletions(-) diff --git a/fs/unionfs/union.h b/fs/unionfs/union.h index 6333488..0476f97 100644 --- a/fs/unionfs/union.h +++ b/fs/unionfs/union.h @@ -142,11 +142,18 @@ struct unionfs_sb_info { * This rwsem is used to make sure that a branch management * operation... * ...

Use vfs helpers and avoid redundant checks performed by the VFS already. Signed-off-by: Erez Zadok <ezk@cs.sunysb.edu> --- fs/unionfs/commonfops.c | 4 --- fs/unionfs/inode.c | 70 +++++++++-------------------------------------- 2 files changed, 13 insertions(+), 61 deletions(-) diff --git a/fs/unionfs/commonfops.c b/fs/unionfs/commonfops.c index 7654bcb..50e5775 100644 --- a/fs/unionfs/commonfops.c +++ b/fs/unionfs/commonfops.c @@ -669,10 +669,6 @@ static long do_ioctl(struct ...

Signed-off-by: Erez Zadok <ezk@cs.sunysb.edu> --- fs/unionfs/inode.c | 9 +++++---- fs/unionfs/rename.c | 2 +- fs/unionfs/subr.c | 8 ++++++-- 3 files changed, 12 insertions(+), 7 deletions(-) diff --git a/fs/unionfs/inode.c b/fs/unionfs/inode.c index 169365c..bcefe45 100644 --- a/fs/unionfs/inode.c +++ b/fs/unionfs/inode.c @@ -465,10 +465,11 @@ static int unionfs_symlink(struct inode *dir, struct dentry *dentry, if (!lower_dentry || IS_ERR(lower_dentry)) { if ...

Signed-off-by: Erez Zadok <ezk@cs.sunysb.edu> --- fs/unionfs/debug.c | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-) diff --git a/fs/unionfs/debug.c b/fs/unionfs/debug.c index 894bf7c..0066ccd 100644 --- a/fs/unionfs/debug.c +++ b/fs/unionfs/debug.c @@ -49,6 +49,9 @@ void __unionfs_check_inode(const struct inode *inode, sb = inode->i_sb; istart = ibstart(inode); iend = ibend(inode); + /* don't check inode if no lower branches */ + if (istart < 0 && iend < ...

Ah, ok. I don't know your projected time frame for this patch series, but if it's for 2.6.25, you might want to look at the set of patches I just sent to lkml to help make files like this easier to create thanks to the great work from Kay Sievers. thanks, greg k-h -

Traditionally sysfs has not been used for "configuration" stuff like this, as that is what configfs is for. Also, a sysfs file that causes an action to happen just by reading the file is not a safe thing to have. Lots of scripts have been known to just walk the whole sysfs tree and open and read everything they can for no real good reason. So you should seriously reconsider this kind of What's wrong with the __ATTR macro? Using a "raw" struct attribute {sigh} Yes, the kernel ...

Hi Steve, I've finished adding the ability for smbd to support up to 16MB writeX calls in the latest git 3.2 tree. To enable, set the parameter : min receivefile size = XXX where XXX is the smallest writeX you want to handle with recvfile. The linux kernel doesn't yet support zerocopy from network to file (ie. splice only works one way currently) so it's emulated in userspace (with a 128k staging buffer) for now. Also it must be an unsigned connection (for obvious reasons). Once ...