btrfs: fix a deadlock in btrfs_scan_one_device()
authorAl Viro <viro@zeniv.linux.org.uk>
Thu, 17 Nov 2011 20:05:22 +0000 (15:05 -0500)
committerAl Viro <viro@zeniv.linux.org.uk>
Mon, 9 Jan 2012 00:33:24 +0000 (19:33 -0500)
commit10f6327b5d26660e06120ba17cff993cb616b1d2
treeee1bade28f62a896db62570589abcb9c8f6f2936
parent6de1d09d9677dce7a04ef485d426821159ab7de9
btrfs: fix a deadlock in btrfs_scan_one_device()

pathname resolution under a global mutex, taken on some paths in ->mount()
is a Bad Idea(tm) - think what happens if said pathname resolution triggers
automount of some btrfs instance and walks into attempt to grab the same
mutex.  Deadlock - we are waiting for daemon to finish walking the path,
daemon is waiting for us to release the mutex...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
fs/btrfs/volumes.c
This page took 0.024534 seconds and 5 git commands to generate.