diff options
author | Deepa Dinamani <deepa.kernel@gmail.com> | 2016-09-14 16:48:03 +0200 |
---|---|---|
committer | Al Viro <viro@zeniv.linux.org.uk> | 2016-09-28 03:06:20 +0200 |
commit | 2554c72edb81c97ae5307613dd0aee1ef8dd13ca (patch) | |
tree | 94e032756f45ea9ced6ddabfcd77e72d7cdfa45d /fs/gfs2 | |
parent | vfs: Add current_time() api (diff) | |
download | linux-2554c72edb81c97ae5307613dd0aee1ef8dd13ca.tar.xz linux-2554c72edb81c97ae5307613dd0aee1ef8dd13ca.zip |
fs: proc: Delete inode time initializations in proc_alloc_inode()
proc uses new_inode_pseudo() to allocate a new inode.
This in turn calls the proc_inode_alloc() callback.
But, at this point, inode is still not initialized
with the super_block pointer which only happens just
before alloc_inode() returns after the call to
inode_init_always().
Also, the inode times are initialized again after the
call to new_inode_pseudo() in proc_inode_alloc().
The assignemet in proc_alloc_inode() is redundant and
also doesn't work after the current_time() api is
changed to take struct inode* instead of
struct *super_block.
This bug was reported after current_time() was used to
assign times in proc_alloc_inode().
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Reported-by: Fengguang Wu <fengguang.wu@intel.com> [0-day test robot]
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Diffstat (limited to 'fs/gfs2')
0 files changed, 0 insertions, 0 deletions