When the exportfs interface was added the expectation was that filesystems
provide an operation to convert from a file handle to an inode/dentry, but it
kept a backwards compat option that still calls into iget.

Calling into iget from non-filesystem code is very bad, because it gives too
little information to filesystem, and simply crashes if the filesystem doesn't
implement the ->read_inode routine.

Fortunately there are only two filesystems left using this fallback: efs and
jfs.  This patch moves a copy of export_iget to each of those to implement the
get_dentry method.

While this is a temporary increase of lines of code in the kernel it allows
for a much cleaner interface and important code restructuring in later
patches.

[akpm@linux-foundation.org: add jfs_get_inode_flags() declaration]
Signed-off-by: NDave Kleikamp <shaggy@linux.vnet.ibm.com>
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NNeil Brown <neilb@suse.de>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Signed-off-by: NDave Kleikamp <shaggy@linux.vnet.ibm.com>

Many struct inode_operations in the kernel can be "const".  Marking them const
moves these to the .rodata section, which avoids false sharing with potential
dirty data.  In addition it'll catch accidental writes at compile time to
these shared resources.
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Yeah, it's about time.
Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

Removed trailing spaces & tabs, and spaces preceding tabs.
Also a couple very minor comment cleanups.
Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>
(cherry picked from f74156539964d7b3d5164fdf8848e6a682f75b97 commit)

I have seen confusing behavior on JFS when I injected many intentional
slab allocation errors. The cp command failed with no disk space error
with enough disk space.

This patch makes:

- change the return value in case slab allocation failures happen
  from -ENOSPC to -ENOMEM

- ialloc() return error code so that the caller can know the reason
  of failures
Signed-off-by: NAkinobu Mita <mita@miraclelinux.com>
Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>
(cherry picked from 2b46f77976f798f3fe800809a1d0ed38763c71c8 commit)

Some filesystems, instead of simply decrementing i_nlink, simply zero it
during an unlink operation.  We need to catch these in addition to the
decrement operations.
Signed-off-by: NDave Hansen <haveblue@us.ibm.com>
Acked-by: NChristoph Hellwig <hch@lst.de>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

This is mostly included for parity with dec_nlink(), where we will have some
more hooks.  This one should stay pretty darn straightforward for now.
Signed-off-by: NDave Hansen <haveblue@us.ibm.com>
Acked-by: NChristoph Hellwig <hch@lst.de>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

When a filesystem decrements i_nlink to zero, it means that a write must be
performed in order to drop the inode from the filesystem.

We're shortly going to have keep filesystems from being remounted r/o between
the time that this i_nlink decrement and that write occurs.

So, add a little helper function to do the decrements.  We'll tie into it in a
bit to note when i_nlink hits zero.
Signed-off-by: NDave Hansen <haveblue@us.ibm.com>
Acked-by: NChristoph Hellwig <hch@lst.de>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

I look at code, and see that
1)locks wasn't release in the opposite order in which they were taken
2)in jfs_rename we lock new_ip, and in "error path" we didn't unlock it
3)I see strange expression: "! !"

May be this worth to fix?
Signed-off-by: NEvgeniy Dushistov <dushistov@mail.ru>
Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

This is a conversion to make the various file_operations structs in fs/
const.  Basically a regexp job, with a few manual fixups

The goal is both to increase correctness (harder to accidentally write to
shared datastructures) and reducing the false sharing of cachelines with
things that get dirty in .data (while .rodata is nicely read only and thus
cache clean)
Signed-off-by: NArjan van de Ven <arjan@infradead.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

ext2 inode attributes with relevance for jfs:

'a' 	EXT2_APPEND_FL       -> append only
'i' 	EXT2_IMMUTABLE_FL    -> immutable file
's' 	EXT2_SECRM_FL	     -> zero file
'u' 	EXT2_UNRM_FL	     -> allow for unrm
'A' 	EXT2_NOATIME_FL      -> no access time
'D' 	EXT2_DIRSYNC_FL      -> dirsync
'S' 	EXT2_SYNC_FL	     -> sync

overview of jfs flags (partially for OS/2)

value	   (OS/2)	Linux	ext2 attrs
------------------------------------------------
0x00010000 IFJOURNAL	-
0x00020000 ISPARSE  	used
0x00040000 INLINEEA 	used
0x00080000 -	    	-	JFS_NOATIME_FL

0x00100000 -	    	-	JFS_DIRSYNC_FL
0x00200000 -	    	-	JFS_SYNC_FL
0x00400000 -	    	-	JFS_SECRM_FL
0x00800000 ISWAPFILE	-	JFS_UNRM_FL

0x01000000 -	    	-	JFS_APPEND_FL
0x02000000 IREADONLY	-	JFS_IMMUTABLE_FL
0x04000000 IHIDDEN  	-	-
0x08000000 ISYSTEM  	-	-

0x10000000 -	    	-
0x20000000 IDIRECTORY	used
0x40000000 IARCHIVE 	-
0x80000000 INEWNAME 	-

the implementation is straight forward, except
for the fact that the attributes have to be mapped
to match with the ext2 ones to avoid a separate
tool for manipulating them (this could be avoided
when using a separate flag field in the on-disk
representation, but the overhead is minimal)

a special jfs_ioctl is added to allow for the new
JFS_IOC_GETFLAGS and JFS_IOC_SETFLAGS calls.

a helper function jfs_set_inode_flags() to transfer
the flags from the on-disk version to the inode

minor changes to allow flag inheritance on inode
creation, as well as a cleanup of the on-disk
flags (including the new ones)

beforementioned helper to map between ext2 and jfs
versions of the new flags ...

the JFS_SECRM_FL and JFS_UNRM_FL are not done yet
and I'm not 100% sure they are worth the effort,
the rest seems to work out of the box ...
Signed-off-by: NHerbert Poetzl <herbert@13thfloor.at>
Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

the conversion was generated via scripts, and the result was validated
automatically via a script as well.

build and boot tested.
Signed-off-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

jfs has never been setting i_ctime or i_mtime when creating either hard
or symbolic links.  I'm surprised nobody had noticed until now.

Thanks to Chris Spiegel for reporting the problem.
Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

This atomically initializes the security xattr when an object is created
Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

Signed-off-by: NDave Kleikamp <shaggy@austin.ibm.com>

Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.

Let it rip!