Commit Graph

144 Commits

Author SHA1 Message Date
Tao Ma c18b812d12 ocfs2: Make transaction extend more efficient.
In ocfs2_extend_rotate_transaction, op_credits is the orignal
credits in the handle and we only want to extend the credits
for the rotation, but the old solution always double it. It
is harmless for some minor operations, but for actions like
reflink we may rotate tree many times and cause the credits
increase dramatically. So this patch try to only increase
the desired credits.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:46 -07:00
Tao Ma 6ae23c5555 ocfs2: CoW refcount tree improvement.
During CoW, if the old extent record is refcounted, we allocate
som new clusters and do CoW. Actually we can have some improvement
here. If the old extent has refcount=1, that means now it is only
used by this file. So we don't need to allocate new clusters, just
remove the refcounted flag and it is OK. We also have to remove
it from the refcount tree while not deleting it.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:36 -07:00
Tao Ma 6f70fa5199 ocfs2: Add CoW support.
This patch try CoW support for a refcounted record.

the whole process will be:
1. Calculate how many clusters we need to CoW and where we start.
   Extents that are not completely encompassed by the write will
   be broken on 1MB boundaries.
2. Do CoW for the clusters with the help of page cache.
3. Change the b-tree structure with the new allocated clusters.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:36 -07:00
Tao Ma bcbbb24a6a ocfs2: Decrement refcount when truncating refcounted extents.
Add 'Decrement refcount for delete' in to the normal truncate
process. So for a refcounted extent record, call refcount rec
decrementation instead of cluster free.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:35 -07:00
Tao Ma 1aa75fea64 ocfs2: Add functions for extents refcounted.
Add function ocfs2_mark_extent_refcounted which can mark
an extent refcounted.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:34 -07:00
Tao Ma 1823cb0b9f ocfs2: Add support of decrementing refcount for delete.
Given a physical cpos and length, decrement the refcount
in the tree. If the refcount for any portion of the extent goes
to zero, that portion is queued for freeing.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:33 -07:00
Tao Ma e2e9f6082b ocfs2: move tree path functions to alloc.h.
Now fs/ocfs2/alloc.c has more than 7000 lines. It contains our
basic b-tree operation. Although we have already make our b-tree
operation generic, the basic structrue ocfs2_path which is used
to iterate one b-tree branch is still static and limited to only
used in alloc.c. As refcount tree need them and I don't want to
add any more b-tree unrelated code to alloc.c, export them out.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:32 -07:00
Tao Ma fe92441595 ocfs2: Add refcount b-tree as a new extent tree.
Add refcount b-tree as a new extent tree so that it can
use the b-tree to store and maniuplate ocfs2_refcount_rec.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:31 -07:00
Tao Ma 555936bfcb ocfs2: Abstract extent split process.
ocfs2_mark_extent_written actually does the following things:
1. check the parameters.
2. initialize the left_path and split_rec.
3. call __ocfs2_mark_extent_written. it will do:
   1) check the flags of unwritten
   2) do the real split work.
The whole process is packed tightly somehow. So this patch
will abstract 2 different functions so that future b-tree
operation can work with it.

1. __ocfs2_split_extent will accept path and split_rec and do
  the real split work.
2. ocfs2_change_extent_flag will accept a new flag and initialize
   path and split_rec.

So now ocfs2_mark_extent_written will do:
1. check the parameters.
2. call ocfs2_change_extent_flag.
   1) initalize the left_path and split_rec.
   2) check whether the new flags conflict with the old one.
   3) call __ocfs2_split_extent to do the split.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:31 -07:00
Tao Ma 853a3a1439 ocfs2: Wrap ocfs2_extent_contig in ocfs2_extent_tree.
Add a new operation eo_ocfs2_extent_contig int the extent tree's
operations vector. So that with the new refcount tree, We want
this so that refcount trees can always return CONTIG_NONE and
prevent extent merging.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
2009-09-22 20:09:30 -07:00
Joel Becker 5e404e9ed1 ocfs2: Pass ocfs2_caching_info into ocfs_init_*_extent_tree().
With this commit, extent tree operations are divorced from inodes and
rely on ocfs2_caching_info.  Phew!

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:13 -07:00
Joel Becker a1cf076ba9 ocfs2: __ocfs2_mark_extent_written() doesn't need struct inode.
We only allow unwritten extents on data, so the toplevel
ocfs2_mark_extent_written() can use an inode all it wants.  But the
subfunction isn't even using the inode argument.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:12 -07:00
Joel Becker f3868d0fa2 ocfs2: Teach ocfs2_replace_extent_rec() to use an extent_tree.
Don't use a struct inode anymore.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:11 -07:00
Joel Becker d231129f44 ocfs2: ocfs2_split_and_insert() no longer needs struct inode.
It already has an extent_tree.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:11 -07:00
Joel Becker dbdcf6a48a ocfs2: ocfs2_remove_extent() no longer needs struct inode.
One more generic btree function that is isolated from struct inode.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:10 -07:00
Joel Becker cbee7e1a6a ocfs2: ocfs2_add_clusters_in_btree() no longer needs struct inode.
One more function that doesn't need a struct inode to pass to its
children.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:09 -07:00
Joel Becker cc79d8c19e ocfs2: ocfs2_insert_extent() no longer needs struct inode.
One more function down, no inode in the entire insert-extent chain.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:09 -07:00
Joel Becker 92ba470c44 ocfs2: Make extent map insertion an extent_tree_operation.
ocfs2_insert_extent() wants to insert a record into the extent map if
it's an inode data extent.  But since many btrees can call that
function, let's make it an op on ocfs2_extent_tree.  Other tree types
can leave it empty.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:08 -07:00
Joel Becker 627961b77e ocfs2: ocfs2_figure_insert_type() no longer needs struct inode.
It's not using it, so remove it from the parameter list.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:08 -07:00
Joel Becker 1ef61b3314 ocfs2: Remove inode from ocfs2_figure_extent_contig().
It already has an ocfs2_extent_tree and doesn't need the inode.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:07 -07:00
Joel Becker a29702914a ocfs2: Swap inode for extent_tree in ocfs2_figure_merge_contig_type().
We don't want struct inode in generic btree operations.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:07 -07:00
Joel Becker b4a176515c ocfs2: ocfs2_extent_contig() only requires the superblock.
Don't pass the inode in.  We don't want it around for generic btree
operations.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:05 -07:00
Joel Becker 3505bec018 ocfs2: ocfs2_do_insert_extent() and ocfs2_insert_path() no longer need an inode.
They aren't using it, so remove it from their parameter lists.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:05 -07:00
Joel Becker c38e52bb1c ocfs2: Give ocfs2_split_record() an extent_tree instead of an inode.
Another on the way to generic btree functions.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:05 -07:00
Joel Becker d562862314 ocfs2: ocfs2_insert_at_leaf() doesn't need struct inode.
Give it an ocfs2_extent_tree and it is happy.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:04 -07:00
Joel Becker 4c911eefca ocfs2: Make truncating the extent map an extent_tree_operation.
ocfs2_remove_extent() wants to truncate the extent map if it's
truncating an inode data extent.  But since many btrees can call that
function, let's make it an op on ocfs2_extent_tree.  Other tree types
can leave it empty.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:03 -07:00
Joel Becker 043beebb6c ocfs2: ocfs2_truncate_rec() doesn't need struct inode.
It's not using it anymore.  Remove it from the parameter list.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:03 -07:00
Joel Becker d401dc12fc ocfs2: ocfs2_grow_branch() and ocfs2_append_rec_to_path() lose struct inode.
ocfs2_grow_branch() not really using it other than to pass it to the
subfunctions ocfs2_shift_tree_depth(), ocfs2_find_branch_target(), and
ocfs2_add_branch().  The first two weren't it either, so they drop the
argument.  ocfs2_add_branch() only passed it to
ocfs2_adjust_rightmost_branch(), which drops the inode argument and uses
the ocfs2_extent_tree as well.

ocfs2_append_rec_to_path() can be take an ocfs2_extent_tree instead of
the inode.  The function ocfs2_adjust_rightmost_records() goes along for
the ride.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:02 -07:00
Joel Becker c495dd24ac ocfs2: ocfs2_try_to_merge_extent() doesn't need struct inode.
It's not using it, so remove it from the parameter list.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:02 -07:00
Joel Becker 4fe82c312a ocfs2: ocfs2_merge_rec_left/right() no longer need struct inode.
Drop it from the parameters - they already have ocfs2_extent_list.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:01 -07:00
Joel Becker 70f18c08b4 ocfs2: ocfs2_rotate_tree_left() no longer needs struct inode.
It already gets ocfs2_extent_tree, so we can just use that.  This chains
to the same modification for ocfs2_remove_rightmost_path() and
ocfs2_rotate_rightmost_leaf_left().

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:08:00 -07:00
Joel Becker e46f74dc35 ocfs2: __ocfs2_rotate_tree_left() doesn't need struct inode.
It already has struct ocfs2_extent_tree, which has the caching info.  So
we don't need to pass it struct inode.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:59 -07:00
Joel Becker 1e2dd63fe0 ocfs2: ocfs2_rotate_subtree_left() doesn't need struct inode.
It already has struct ocfs2_extent_tree, which has the caching info.  So
we don't need to pass it struct inode.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:59 -07:00
Joel Becker 09106bae05 ocfs2: ocfs2_update_edge_lengths() doesn't need struct inode.
Pass in the extent tree, which is all we need.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:58 -07:00
Joel Becker 1bbf0b8d60 ocfs2: ocfs2_rotate_tree_right() doesn't need struct inode.
We don't need struct inode in ocfs2_rotate_tree_right() anymore.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:58 -07:00
Joel Becker 6136ca5f5f ocfs2: Drop struct inode from ocfs2_extent_tree_operations.
We can get to the inode from the caching information.  Other parent
types don't need it.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:57 -07:00
Joel Becker 7dc0280567 ocfs2: Pass ocfs2_extent_tree to ocfs2_get_subtree_root()
Get rid of the inode argument.  Use extent_tree instead.  This means a
few more functions have to pass an extent_tree around.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:55 -07:00
Joel Becker 5c601aba8c ocfs2: Get inode out of ocfs2_rotate_subtree_root_right().
Pass the ocfs2_extent_list down through ocfs2_rotate_tree_right() and
get rid of struct inode in ocfs2_rotate_subtree_root_right().

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:55 -07:00
Joel Becker 4619c73e7c ocfs2: ocfs2_complete_edge_insert() doesn't need struct inode at all.
Completely unused argument.  Get rid of it.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:54 -07:00
Joel Becker 6641b0ce32 ocfs2: Pass ocfs2_extent_tree to ocfs2_unlink_path()
ocfs2_unlink_path() doesn't need struct inode, so let's pass it struct
ocfs2_extent_tree.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:54 -07:00
Joel Becker 42a5a7a9a5 ocfs2: ocfs2_create_new_meta_bhs() doesn't need struct inode.
Pass struct ocfs2_extent_tree into ocfs2_create_new_meta_bhs().  It no
longer needs struct inode or ocfs2_super.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:53 -07:00
Joel Becker facdb77f54 ocfs2: ocfs2_find_path() only needs the caching info
ocfs2_find_path and ocfs2_find_leaf() walk our btrees, reading extent
blocks.  They need struct ocfs2_caching_info for that, but not struct
inode.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:53 -07:00
Joel Becker 3d03a305de ocfs2: Pass ocfs2_caching_info to ocfs2_read_extent_block().
extent blocks belong to btrees on more than just inodes, so we want to
pass the ocfs2_caching_info structure directly to
ocfs2_read_extent_block().  A number of places in alloc.c can now drop
struct inode from their argument list.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:52 -07:00
Joel Becker d9a0a1f83b ocfs2: Store the ocfs2_caching_info on ocfs2_extent_tree.
What do we cache?  Metadata blocks.  What are most of our non-inode metadata
blocks?  Extent blocks for our btrees.  struct ocfs2_extent_tree is the
main structure for managing those.  So let's store the associated
ocfs2_caching_info there.

This means that ocfs2_et_root_journal_access() doesn't need struct inode
anymore, and any place that has an et can refer to et->et_ci instead of
INODE_CACHE(inode).

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:51 -07:00
Joel Becker 0cf2f7632b ocfs2: Pass struct ocfs2_caching_info to the journal functions.
The next step in divorcing metadata I/O management from struct inode is
to pass struct ocfs2_caching_info to the journal functions.  Thus the
journal locks a metadata cache with the cache io_lock function.  It also
can compare ci_last_trans and ci_created_trans directly.

This is a large patch because of all the places we change
ocfs2_journal_access..(handle, inode, ...) to
ocfs2_journal_access..(handle, INODE_CACHE(inode), ...).

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:50 -07:00
Joel Becker 8cb471e8f8 ocfs2: Take the inode out of the metadata read/write paths.
We are really passing the inode into the ocfs2_read/write_blocks()
functions to get at the metadata cache.  This commit passes the cache
directly into the metadata block functions, divorcing them from the
inode.

Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-09-04 16:07:48 -07:00
Tao Ma 60e2ec4866 ocfs2: release the buffer head in ocfs2_do_truncate.
In ocfs2_do_truncate, we forget to release last_eb_bh which
will cause memleak. So call brelse in the end.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-08-17 12:50:35 -07:00
Tao Ma 82e12644cf ocfs2: Use ocfs2_rec_clusters in ocfs2_adjust_adjacent_records.
In ocfs2_adjust_adjacent_records, we will adjust adjacent records
according to the extent_list in the lower level. But actually
the lower level tree will either be a leaf or a branch. If we only
use ocfs2_is_empty_extent we will meet with some problem if the lower
tree is a branch (tree_depth > 1). So use !ocfs2_rec_clusters instead.
And actually only the leaf record can have holes. So add a BUG_ON
for non-leaf branch.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-07-23 10:58:46 -07:00
Tao Ma 3c5e10683e ocfs2: Add extra credits and access the modified bh in update_edge_lengths.
In normal tree rotation left process, we will never touch the tree
branch above subtree_index and ocfs2_extend_rotate_transaction doesn't
reserve the credits for them either.

But when we want to delete the rightmost extent block, we have to update
the rightmost records for all the rightmost branch(See
ocfs2_update_edge_lengths), so we have to allocate extra credits for them.
What's more, we have to access them also.

Signed-off-by: Tao Ma <tao.ma@oracle.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-07-21 14:41:54 -07:00
Tao Ma 6b791bcc8b ocfs2: Adjust rightmost path in ocfs2_add_branch.
In ocfs2_add_branch, we use the rightmost rec of the leaf extent block
to generate the e_cpos for the newly added branch. In the most case, it
is OK but if the parent extent block's rightmost rec covers more clusters
than the leaf does, it will cause kernel panic if we insert some clusters
in it. The message is something like:
(7445,1):ocfs2_insert_at_leaf:3775 ERROR: bug expression:
le16_to_cpu(el->l_next_free_rec) >= le16_to_cpu(el->l_count)
(7445,1):ocfs2_insert_at_leaf:3775 ERROR: inode 66053, depth 0, count 28,
next free 28, rec.cpos 270, rec.clusters 1, insert.cpos 275, insert.clusters 1
 [<fa7ad565>] ? ocfs2_do_insert_extent+0xb58/0xda0 [ocfs2]
 [<fa7b08f2>] ? ocfs2_insert_extent+0x5bd/0x6ba [ocfs2]
 [<fa7b1b8b>] ? ocfs2_add_clusters_in_btree+0x37f/0x564 [ocfs2]
...

The panic can be easily reproduced by the following small test case
(with bs=512, cs=4K, and I remove all the error handling so that it looks
clear enough for reading).

int main(int argc, char **argv)
{
	int fd, i;
	char buf[5] = "test";

	fd = open(argv[1], O_RDWR|O_CREAT);

	for (i = 0; i < 30; i++) {
		lseek(fd, 40960 * i, SEEK_SET);
		write(fd, buf, 5);
	}

	ftruncate(fd, 1146880);

	lseek(fd, 1126400, SEEK_SET);
	write(fd, buf, 5);

	close(fd);

	return 0;
}

The reason of the panic is that:
the 30 writes and the ftruncate makes the file's extent list looks like:

	Tree Depth: 1   Count: 19   Next Free Rec: 1
	## Offset        Clusters       Block#
	0  0             280            86183
	SubAlloc Bit: 7   SubAlloc Slot: 0
	Blknum: 86183   Next Leaf: 0
	CRC32: 00000000   ECC: 0000
	Tree Depth: 0   Count: 28   Next Free Rec: 28
	## Offset        Clusters       Block#          Flags
	0  0             1              143368          0x0
	1  10            1              143376          0x0
	...
	26 260           1              143576          0x0
	27 270           1              143584          0x0

Now another write at 1126400(275 cluster) whiich will write at the gap
between 271 and 280 will trigger ocfs2_add_branch, but the result after
the function looks like:
	Tree Depth: 1   Count: 19   Next Free Rec: 2
	## Offset        Clusters       Block#
	0  0             280            86183
	1  271           0             143592
So the extent record is intersected and make the following operation bug out.

This patch just try to remove the gap before we add the new branch, so that
the root(branch) rightmost rec will cover the same right position. So in the
above case, before adding branch the tree will be changed to
	Tree Depth: 1   Count: 19   Next Free Rec: 1
	## Offset        Clusters       Block#
	0  0             271            86183
	SubAlloc Bit: 7   SubAlloc Slot: 0
	Blknum: 86183   Next Leaf: 0
	CRC32: 00000000   ECC: 0000
	Tree Depth: 0   Count: 28   Next Free Rec: 28
	## Offset        Clusters       Block#          Flags
	0  0             1              143368          0x0
	1  10            1              143376          0x0
	...
	26 260           1              143576          0x0
	27 270           1              143584          0x0
And after branch add, the tree looks like
	Tree Depth: 1   Count: 19   Next Free Rec: 2
	## Offset        Clusters       Block#
	0  0             271            86183
	1  271           0             143592

Signed-off-by: Tao Ma <tao.ma@oracle.com>
Acked-by: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
2009-06-15 14:49:43 -07:00