[PATCH] Btrfs: fix dentry->d_parent abuses V2

2010-11-08 Thread Josef Bacik
There are lots of places where we do dentry->d_parent->d_inode without holding the dentry->d_lock. This could cause problems with rename. So instead use dget_parent where we can, or in some cases we don't even need to use dentry->d_parent->d_inode since we get the inode of the dir passed to us fr

Re: [PATCH] Btrfs: fix dentry->d_parent abuses

2010-10-28 Thread Ian Kent
On Thu, 2010-10-28 at 02:43 -0400, Josef Bacik wrote: > On Thu, Oct 28, 2010 at 11:42:04AM +0800, Ian Kent wrote: > > On Wed, 2010-10-27 at 10:39 -0400, Josef Bacik wrote: > > > There are lots of places where we do dentry->d_parent->d_inode without > > > holding > > > the dentry->d_lock. This cou

Re: [PATCH] Btrfs: fix dentry->d_parent abuses

2010-10-27 Thread Josef Bacik
On Thu, Oct 28, 2010 at 11:42:04AM +0800, Ian Kent wrote: > On Wed, 2010-10-27 at 10:39 -0400, Josef Bacik wrote: > > There are lots of places where we do dentry->d_parent->d_inode without > > holding > > the dentry->d_lock. This could cause problems with rename. So instead use > > dget_parent w

Re: [PATCH] Btrfs: fix dentry->d_parent abuses

2010-10-27 Thread Ian Kent
On Wed, 2010-10-27 at 10:39 -0400, Josef Bacik wrote: > There are lots of places where we do dentry->d_parent->d_inode without holding > the dentry->d_lock. This could cause problems with rename. So instead use > dget_parent where we can, or in some cases we don't even need to use > dentry->d_par