GIT 788d99bcd70ad5c3aa22468040c47460ed38b06d git+ssh://master.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw.git commit Author: Marcin Slusarz Date: Wed Feb 13 00:06:10 2008 +0100 [GFS2] be*_add_cpu conversion replace all: big_endian_variable = cpu_to_beX(beX_to_cpu(big_endian_variable) + expression_in_cpu_byteorder); with: beX_add_cpu(&big_endian_variable, expression_in_cpu_byteorder); generated with semantic patch Signed-off-by: Marcin Slusarz Signed-off-by: Steven Whitehouse commit 10fa64adf069cad5f660266c94886fd9cb496430 Author: Steven Whitehouse Date: Tue Feb 12 15:28:21 2008 +0000 [GFS2] Fix bug where we called drop_bh incorrectly As a result of an earlier patch, drop_bh was being called in cases when it shouldn't have been. Since we never have a gh in the drop case and we always have a gh in the promote case, we can use that extra information to tell which case has been seen. Signed-off-by: Steven Whitehouse Cc: Bob Peterson commit dc3cce3c1cbab73146759a8b4bf39951644e86fb Author: Steven Whitehouse Date: Tue Feb 12 14:48:39 2008 +0000 [GFS2] Get inode buffer only once per block map call In the case that we needed to grow the height of the metadata tree we were looking up the inode buffer and then brelse()ing it despite the fact that it is needed later in the block map process. This patch ensures that we look up the inode's buffer once and only once during the block map process. Signed-off-by: Steven Whitehouse commit cb23c9e10ac5af03176b252d41a31eb160bfc921 Author: Steven Whitehouse Date: Tue Feb 12 14:17:27 2008 +0000 [GFS2] Eliminate (almost) duplicate field from gfs2_inode The blocks counter is almost a duplicate of the i_blocks field in the VFS inode. The only difference is that i_blocks can be only 32bits long for 32bit arch without large single file support. Since GFS2 doesn't handle the non-large single file case (for 32 bit anyway) this adds a new config dependency on 64BIT || LSF. This has always been the case, however we've never explicitly said so before. Even if we do add support for the non-LSF case, we will still not require this field to be duplicated since we will not be able to access oversized files anyway. So the net result of all this is that we shave 8 bytes from a gfs2_inode and get our config deps correct. Signed-off-by: Steven Whitehouse commit 07a33183e1adba511eacd85ab95adc40be5cf1bf Author: Steven Whitehouse Date: Fri Feb 8 13:18:11 2008 +0000 [GFS2] Add a function to interate over an extent This adds a function (currently the only use is during mapping of already allocated blocks, but watch this space) which iterates over a number of pointers in a block and returns the extent length. If the initial pointer is 0 (i.e. unallocated) it will return the number of unallocated blocks in the extent. If the initial pointer is allocated, then it returns the number of contiguously allocated blocks in the extent. Signed-off-by: Steven Whitehouse commit 1ce7998a0f908fa14ead1852e7fb56b49cf8faec Author: Steven Whitehouse Date: Tue Feb 12 12:14:59 2008 +0000 [GFS2] The case of the missing asterisk A dereference was forgotten. This adds it back correctly. Signed-off-by: Steven Whitehouse commit 63e3879c58d60aa8c8c7a2f185c1c5234a56f264 Author: Steven Whitehouse Date: Wed Feb 6 10:11:15 2008 +0000 [GFS2] Add extent allocation to block allocator Rather than having to allocate a single block at a time, this patch allows the block allocator to allocate an extent. Since there is no difference (so far as the block allocator is concerned) between data blocks and indirect blocks, it is posible to allocate a single extent and for the caller to unrevoke just the blocks required for indirect blocks. Currently the only bit of GFS2 to make use of this feature is the build height function. The intention is that gfs2_block_map will be changed to make use of this feature in future patches. Signed-off-by: Steven Whitehouse commit c7a7e352d30420a4e2c2436053db6d81ff9259a4 Author: Steven Whitehouse Date: Fri Feb 1 14:52:30 2008 +0000 [GFS2] Merge gfs2_alloc_meta and gfs2_alloc_data Thanks to the preceeding patches, the only difference between these two functions is their name. We can thus merge them and call the new function gfs2_alloc_block to reflect the fact that it can allocate either kind of block. Signed-off-by: Steven Whitehouse commit 83af44420543c75834402c2b71b33995fb9a08af Author: Steven Whitehouse Date: Fri Feb 1 13:16:55 2008 +0000 [GFS2] Update gfs2_trans_add_unrevoke to accept extents By adding an extra argument to gfs2_trans_add_unrevoke we can now specify an extent length of blocks to unrevoke. This means that we only need to make one pass through the list for each extent rather than each block. Currently the only extent length which is used is 1, but that will change in the future. Also gfs2_trans_add_unrevoke is removed from gfs2_alloc_meta since its the only difference between this and gfs2_alloc_data which is left. This will allow a future patch to merge these two functions into one (i.e. one call to allocate both data and metadata in a single extent in the future). Signed-off-by: Steven Whitehouse commit 1f4c4ce550d3f0d62ed9499cce6f5138a193813a Author: Steven Whitehouse Date: Fri Feb 1 10:34:15 2008 +0000 [GFS2] Merge the rd_last_alloc_meta and rd_last_alloc_data fields We don't need to keep track of when we last allocated data and metadata separately since the only thing thats important when searching for a free block is whether its free or not, which is independent from what type of block it is. Signed-off-by: Steven Whitehouse commit 311022435c46af87b33934c7ae878b39954dedc8 Author: Steven Whitehouse Date: Wed Feb 6 09:25:45 2008 +0000 [GFS2] Reduce inode size by merging fields There were three fields being used to keep track of the location of the most recently allocated block for each inode. These have been merged into a single field in order to better keep the data and metadata for an inode close on disk, and also to reduce the space required for storage. Signed-off-by: Steven Whitehouse commit c2723d463c97f4d29a8d9ba2ef7c082c076e524f Author: Bob Peterson Date: Tue Feb 5 17:11:40 2008 -0600 [GFS2] Remove unused counters This is kind of trivial in the greater scheme of things, but this removes three counters that AFAICT are never used. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit ee32b69b0bebfef680bb81eda3e0096c04ae1888 Author: Steven Whitehouse Date: Fri Feb 1 09:23:44 2008 +0000 [GFS2] Shrink & rename di_depth This patch forms a pair with the previous patch which shrunk di_height. Like that patch di_depth is renamed i_depth and moved into struct gfs2_inode directly. Also the field goes from 16 bits to 8 bits since it is also limited to a max value which is rather small (17 in this case). In addition we also now validate the field against this maximum value when its read in. Signed-off-by: Steven Whitehouse commit 12311aee9adea51341e82a888e87770c62151c8d Author: Bob Peterson Date: Thu Jan 31 10:31:39 2008 -0600 [GFS2] Remove rgrp and glock version numbers This patch further reduces GFS2's memory requirements by eliminating the 64-bit version number fields in lieu of a couple bits. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit fb85bfc1ce60d64b8d3eda7e2804bb430ecb50e4 Author: Steven Whitehouse Date: Wed Jan 30 15:34:04 2008 +0000 [GFS2] Remove lm.[ch] and distribute content The functions in lm.c were just wrappers which were mostly only used in one other file. By moving the functions to the files where they are being used, they can be marked static and also this will usually result in them being inlined since they are often only used from one point in the code. A couple of really trivial functions have been inlined by hand into the function which called them as it makes the code clearer to do that. We also gain from one fewer function call in the glock lock and unlock paths. Signed-off-by: Steven Whitehouse commit d1c792b30d2d09683b4fdcc8b7035593ab98dcec Author: Bob Peterson Date: Tue Jan 29 13:56:15 2008 -0600 [GFS2] Eliminate gl_req_bh This patch further reduces the memory needs of GFS2 by eliminating the gl_req_bh variable from struct gfs2_glock. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit 1e3692ba918945070d50cdd895c1639212f86b53 Author: Steven Whitehouse Date: Tue Jan 29 13:30:20 2008 +0000 [GFS2] Add consts to various bits of rgrp.c There are a couple of routines which scan bitmaps where we can mark the bitmaps const, plus a couple of call sites that can be updated too. Signed-off-by: Steven Whitehouse commit 266d8a9c361935bfae1379916f1ce2058dd215de Author: Steven Whitehouse Date: Tue Jan 29 09:12:55 2008 +0000 [GFS2] Introduce array of buffers to struct metapath The reason for doing this is to allow all the block mapping code to share the same array. As a result we can remove two arguments from lookup_metapath since they are now returned via the array. We also add a function to drop all refs to buffer heads when we are done with the metapath. The build_height function shares the struct metapath, but currently still frees its own buffers, and this will change in a future patch. Signed-off-by: Steven Whitehouse commit ed6522c4d9b90df796f71e49ac43b548e8582d4b Author: Steven Whitehouse Date: Mon Jan 28 15:10:29 2008 +0000 [GFS2] Move part of gfs2_block_map into a separate function This is required to enable future changes to the block mapping code. Signed-off-by: Steven Whitehouse commit 17e379fb31bea6051d37e2af2dc4b67cb65fd718 Author: Bob Peterson Date: Mon Jan 28 22:31:39 2008 -0600 [GFS2] Get rid of gl_waiters2 This patch reduces memory by replacing the int variable gl_waiters2 by a single bit in the gl_flags. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit ed03de5dda20f3b72c8696cf9cca818be9b10245 Author: Bob Peterson Date: Mon Jan 28 18:38:07 2008 -0600 [GFS2] Combine rg_flags and rd_flags This patch reduces the memory required by GFS2 by combining the rd_flags and rg_flags (in core only). Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit 16ad345b87a0dd98cc3eadbfed5f7a1f1cfa47f4 Author: Bob Peterson Date: Mon Jan 28 17:20:26 2008 -0600 [GFS2] Allocate gfs2_rgrpd from slab memory This patch moves the gfs2_rgrpd structure to its own slab memory. This makes it easier to control and monitor, and yields less memory fragmentation. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit a0539eaea41753126ab856f53221ce16cdab08bd Author: Bob Peterson Date: Mon Jan 28 16:35:13 2008 -0600 [GFS2] Plug an unlikely leak Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit 37bcd9c8dfe64ac1f24d048629a11a7bfea6b67e Author: Adrian Bunk Date: Tue Jan 29 00:11:34 2008 +0200 [GFS2] make gfs2_glock_hold() static gfs2_glock_hold() can now become static. Signed-off-by: Adrian Bunk Signed-off-by: Steven Whitehouse commit 8e3f06d45e9b5a4105bb256d45b62c778b66229a Author: Bob Peterson Date: Mon Jan 28 14:54:16 2008 -0600 [GFS2] Only wake the reclaim daemon if we need to This patch only wakes up the glock reclaim daemon if there is actually something to be reclaimed. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit f24a9a7a5480816c6d5953fffdcec38cd6658cef Author: Bob Peterson Date: Mon Jan 28 11:24:35 2008 -0600 [GFS2] Misc fixups This patch contains two small fixups that didn't fit elsewhere. They are: (1) get rid of temp variable in find_metapath. (2) Remove vestigial "ret" variable from gfs2_writepage_common. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit 9a8f40316da493bbe030a5eda88e6fcf98a6ae80 Author: Bob Peterson Date: Mon Jan 28 11:20:10 2008 -0600 [GFS2] Only do lo_incore_commit once This patch is performance related. When we're doing a log flush, I noticed we were calling buf_lo_incore_commit twice: once for data bufs and once for metadata bufs. Since this is the same function and does the same thing in both cases, there should be no reason to call it twice. Since we only need to call it once, we can also make it faster by removing it from the generic "lops" code and making it a stand-along static function. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit 6a79cadc487f06717402fa1bf053346fc0dfe2a1 Author: Bob Peterson Date: Mon Jan 28 11:15:57 2008 -0600 [GFS2] Fix debug inode printing I noticed that the latest change to i_height got rid of the value from the inode dump. This patch adds it back. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit 81349107bed376e1059248d6b553dfe3b1db0430 Author: Bob Peterson Date: Mon Jan 28 11:13:02 2008 -0600 [GFS2] Get rid of unneeded parameter in gfs2_rlist_alloc This patch removed the unnecessary parameter from function gfs2_rlist_alloc. The parameter was always passed in as 0. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse commit c4a3dc17a7103d386dc5f032491429147dd9447c Author: Steven Whitehouse Date: Mon Jan 28 10:37:35 2008 +0000 [GFS2] Streamline indirect pointer tree height calculation This patch improves the calculation of the tree height in order to reduce the number of operations which are carried out on each call to gfs2_block_map. In the common case, we now make a single comparison, rather than calculating the required tree height from scratch each time. Also in the case that the tree does need some extra height, we start from the current height rather from zero when we work out what the new height ought to be. In addition the di_height field is moved into the inode proper and reduced in size to a u8 since the value must be between 0 and GFS2_MAX_META_HEIGHT (10). Signed-off-by: Steven Whitehouse commit c0ee89384008cc9ab8fd5e6226c27f9d5543081d Author: Steven Whitehouse Date: Mon Jan 28 08:47:38 2008 +0000 [GFS2] Speed up gfs2_write_alloc_required, deprecate gfs2_extent_map This patch removes the call to gfs2_extent_map from gfs2_write_alloc_required, instead we call gfs2_block_map directly. This results in fewer overall calls to gfs2_block_map in the multi-block case. Also, gfs2_extent_map is marked as deprecated so that people know that its going away as soon as all the callers have been converted. Signed-off-by: Steven Whitehouse fs/gfs2/Kconfig | 2 +- fs/gfs2/Makefile | 2 +- fs/gfs2/bmap.c | 379 +++++++++++++++++++++++++------------------------ fs/gfs2/dir.c | 62 ++++----- fs/gfs2/eattr.c | 42 ++---- fs/gfs2/glock.c | 159 ++++++++++++--------- fs/gfs2/glock.h | 1 - fs/gfs2/glops.c | 10 +- fs/gfs2/incore.h | 39 ++--- fs/gfs2/inode.c | 58 ++++---- fs/gfs2/inode.h | 22 +++- fs/gfs2/lm.c | 210 --------------------------- fs/gfs2/lm.h | 42 ------ fs/gfs2/log.c | 17 ++- fs/gfs2/lops.c | 21 +--- fs/gfs2/lops.h | 11 +-- fs/gfs2/main.c | 10 ++ fs/gfs2/ops_address.c | 4 +- fs/gfs2/ops_file.c | 31 ++++- fs/gfs2/ops_fstype.c | 75 ++++++++++- fs/gfs2/ops_inode.c | 5 +- fs/gfs2/ops_super.c | 1 - fs/gfs2/recovery.c | 11 ++- fs/gfs2/rgrp.c | 293 ++++++++++++++++++-------------------- fs/gfs2/rgrp.h | 8 +- fs/gfs2/super.c | 2 + fs/gfs2/super.h | 1 + fs/gfs2/sys.c | 7 - fs/gfs2/trans.c | 25 ++-- fs/gfs2/trans.h | 2 +- fs/gfs2/util.c | 24 +++- fs/gfs2/util.h | 2 + 32 files changed, 729 insertions(+), 849 deletions(-) diff --git a/fs/gfs2/Kconfig b/fs/gfs2/Kconfig index de8e64c..d147b53 100644 --- a/fs/gfs2/Kconfig +++ b/fs/gfs2/Kconfig @@ -1,6 +1,6 @@ config GFS2_FS tristate "GFS2 file system support" - depends on EXPERIMENTAL + depends on EXPERIMENTAL && (64BIT || LSF) select FS_POSIX_ACL select CRC32 help diff --git a/fs/gfs2/Makefile b/fs/gfs2/Makefile index 8fff110..e2350df 100644 --- a/fs/gfs2/Makefile +++ b/fs/gfs2/Makefile @@ -1,6 +1,6 @@ obj-$(CONFIG_GFS2_FS) += gfs2.o gfs2-y := acl.o bmap.o daemon.o dir.o eaops.o eattr.o glock.o \ - glops.o inode.o lm.o log.o lops.o locking.o main.o meta_io.o \ + glops.o inode.o log.o lops.o locking.o main.o meta_io.o \ mount.o ops_address.o ops_dentry.o ops_export.o ops_file.o \ ops_fstype.o ops_inode.o ops_super.o quota.o \ recovery.o rgrp.o super.o sys.o trans.o util.o diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index e9456eb..f1f38ca 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -33,6 +33,7 @@ * keep it small. */ struct metapath { + struct buffer_head *mp_bh[GFS2_MAX_META_HEIGHT]; __u16 mp_list[GFS2_MAX_META_HEIGHT]; }; @@ -135,9 +136,10 @@ int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page) /* Get a free block, fill it with the stuffed data, and write it out to disk */ + unsigned int n = 1; + block = gfs2_alloc_block(ip, &n); if (isdir) { - block = gfs2_alloc_meta(ip); - + gfs2_trans_add_unrevoke(GFS2_SB(&ip->i_inode), block, 1); error = gfs2_dir_get_new_buffer(ip, block, &bh); if (error) goto out_brelse; @@ -145,8 +147,6 @@ int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page) dibh, sizeof(struct gfs2_dinode)); brelse(bh); } else { - block = gfs2_alloc_data(ip); - error = gfs2_unstuffer_page(ip, dibh, block, page); if (error) goto out_brelse; @@ -161,12 +161,11 @@ int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page) if (ip->i_di.di_size) { *(__be64 *)(di + 1) = cpu_to_be64(block); - ip->i_di.di_blocks++; - gfs2_set_inode_blocks(&ip->i_inode); - di->di_blocks = cpu_to_be64(ip->i_di.di_blocks); + gfs2_add_inode_blocks(&ip->i_inode, 1); + di->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); } - ip->i_di.di_height = 1; + ip->i_height = 1; di->di_height = cpu_to_be16(1); out_brelse: @@ -177,43 +176,6 @@ out: } /** - * calc_tree_height - Calculate the height of a metadata tree - * @ip: The GFS2 inode - * @size: The proposed size of the file - * - * Work out how tall a metadata tree needs to be in order to accommodate a - * file of a particular size. If size is less than the current size of - * the inode, then the current size of the inode is used instead of the - * supplied one. - * - * Returns: the height the tree should be - */ - -static unsigned int calc_tree_height(struct gfs2_inode *ip, u64 size) -{ - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - u64 *arr; - unsigned int max, height; - - if (ip->i_di.di_size > size) - size = ip->i_di.di_size; - - if (gfs2_is_dir(ip)) { - arr = sdp->sd_jheightsize; - max = sdp->sd_max_jheight; - } else { - arr = sdp->sd_heightsize; - max = sdp->sd_max_height; - } - - for (height = 0; height < max; height++) - if (arr[height] >= size) - break; - - return height; -} - -/** * build_height - Build a metadata tree of the requested height * @ip: The GFS2 inode * @height: The height to build to @@ -222,61 +184,58 @@ static unsigned int calc_tree_height(struct gfs2_inode *ip, u64 size) * Returns: errno */ -static int build_height(struct inode *inode, unsigned height) +static int build_height(struct inode *inode, struct metapath *mp, unsigned height) { struct gfs2_inode *ip = GFS2_I(inode); - unsigned new_height = height - ip->i_di.di_height; - struct buffer_head *dibh; - struct buffer_head *blocks[GFS2_MAX_META_HEIGHT]; + unsigned new_height = height - ip->i_height; + struct buffer_head *dibh = mp->mp_bh[0]; struct gfs2_dinode *di; - int error; __be64 *bp; u64 bn; - unsigned n; + unsigned n, i = 0; - if (height <= ip->i_di.di_height) - return 0; + BUG_ON(height <= ip->i_height); - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - return error; - - for(n = 0; n < new_height; n++) { - bn = gfs2_alloc_meta(ip); - blocks[n] = gfs2_meta_new(ip->i_gl, bn); - gfs2_trans_add_bh(ip->i_gl, blocks[n], 1); - } + do { + n = new_height - i; + bn = gfs2_alloc_block(ip, &n); + gfs2_trans_add_unrevoke(GFS2_SB(inode), bn, n); + do { + mp->mp_bh[i + 1] = gfs2_meta_new(ip->i_gl, bn++); + gfs2_trans_add_bh(ip->i_gl, mp->mp_bh[i + 1], 1); + i++; + } while(i < n); + } while(i < new_height); n = 0; - bn = blocks[0]->b_blocknr; + bn = mp->mp_bh[1]->b_blocknr; if (new_height > 1) { for(; n < new_height-1; n++) { - gfs2_metatype_set(blocks[n], GFS2_METATYPE_IN, + gfs2_metatype_set(mp->mp_bh[n + 1], GFS2_METATYPE_IN, GFS2_FORMAT_IN); - gfs2_buffer_clear_tail(blocks[n], + gfs2_buffer_clear_tail(mp->mp_bh[n + 1], sizeof(struct gfs2_meta_header)); - bp = (__be64 *)(blocks[n]->b_data + + bp = (__be64 *)(mp->mp_bh[n + 1]->b_data + sizeof(struct gfs2_meta_header)); - *bp = cpu_to_be64(blocks[n+1]->b_blocknr); - brelse(blocks[n]); - blocks[n] = NULL; + *bp = cpu_to_be64(mp->mp_bh[n+2]->b_blocknr); + brelse(mp->mp_bh[n+1]); + mp->mp_bh[n+1] = NULL; } } - gfs2_metatype_set(blocks[n], GFS2_METATYPE_IN, GFS2_FORMAT_IN); - gfs2_buffer_copy_tail(blocks[n], sizeof(struct gfs2_meta_header), + gfs2_metatype_set(mp->mp_bh[n+1], GFS2_METATYPE_IN, GFS2_FORMAT_IN); + gfs2_buffer_copy_tail(mp->mp_bh[n+1], sizeof(struct gfs2_meta_header), dibh, sizeof(struct gfs2_dinode)); - brelse(blocks[n]); + brelse(mp->mp_bh[n+1]); + mp->mp_bh[n+1] = NULL; gfs2_trans_add_bh(ip->i_gl, dibh, 1); di = (struct gfs2_dinode *)dibh->b_data; gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); *(__be64 *)(di + 1) = cpu_to_be64(bn); - ip->i_di.di_height += new_height; - ip->i_di.di_blocks += new_height; - gfs2_set_inode_blocks(&ip->i_inode); - di->di_height = cpu_to_be16(ip->i_di.di_height); - di->di_blocks = cpu_to_be64(ip->i_di.di_blocks); - brelse(dibh); - return error; + ip->i_height += new_height; + gfs2_add_inode_blocks(&ip->i_inode, new_height); + di->di_height = cpu_to_be16(ip->i_height); + di->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); + return 0; } /** @@ -342,17 +301,15 @@ static void find_metapath(struct gfs2_inode *ip, u64 block, struct metapath *mp) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - u64 b = block; unsigned int i; - for (i = ip->i_di.di_height; i--;) - mp->mp_list[i] = do_div(b, sdp->sd_inptrs); + for (i = ip->i_height; i--;) + mp->mp_list[i] = do_div(block, sdp->sd_inptrs); } /** * metapointer - Return pointer to start of metadata in a buffer - * @bh: The buffer * @height: The metadata height (0 = dinode) * @mp: The metapath * @@ -361,9 +318,10 @@ static void find_metapath(struct gfs2_inode *ip, u64 block, * metadata tree. */ -static inline __be64 *metapointer(struct buffer_head *bh, int *boundary, - unsigned int height, const struct metapath *mp) +static inline __be64 *metapointer(int *boundary, unsigned int height, + const struct metapath *mp) { + struct buffer_head *bh = mp->mp_bh[height]; unsigned int head_size = (height > 0) ? sizeof(struct gfs2_meta_header) : sizeof(struct gfs2_dinode); __be64 *ptr; @@ -377,7 +335,6 @@ static inline __be64 *metapointer(struct buffer_head *bh, int *boundary, /** * lookup_block - Get the next metadata block in metadata tree * @ip: The GFS2 inode - * @bh: Buffer containing the pointers to metadata blocks * @height: The height of the tree (0 = dinode) * @mp: The metapath * @create: Non-zero if we may create a new meatdata block @@ -390,12 +347,13 @@ static inline __be64 *metapointer(struct buffer_head *bh, int *boundary, * */ -static int lookup_block(struct gfs2_inode *ip, struct buffer_head *bh, - unsigned int height, struct metapath *mp, int create, +static int lookup_block(struct gfs2_inode *ip, unsigned int height, + struct metapath *mp, int create, int *new, u64 *block) { int boundary; - __be64 *ptr = metapointer(bh, &boundary, height, mp); + __be64 *ptr = metapointer(&boundary, height, mp); + unsigned int n = 1; if (*ptr) { *block = be64_to_cpu(*ptr); @@ -407,21 +365,86 @@ static int lookup_block(struct gfs2_inode *ip, struct buffer_head *bh, if (!create) return 0; - if (height == ip->i_di.di_height - 1 && !gfs2_is_dir(ip)) - *block = gfs2_alloc_data(ip); - else - *block = gfs2_alloc_meta(ip); + *block = gfs2_alloc_block(ip, &n); + if (height != ip->i_height - 1 || gfs2_is_dir(ip)) + gfs2_trans_add_unrevoke(GFS2_SB(&ip->i_inode), *block, 1); - gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_trans_add_bh(ip->i_gl, mp->mp_bh[height], 1); *ptr = cpu_to_be64(*block); - ip->i_di.di_blocks++; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, 1); *new = 1; return 0; } +static int lookup_metapath(struct inode *inode, struct metapath *mp, + int create, int *new, u64 *dblock) +{ + struct buffer_head *bh; + struct gfs2_inode *ip = GFS2_I(inode); + unsigned int end_of_metadata = ip->i_height - 1; + unsigned int x; + int ret; + + for (x = 0; x < end_of_metadata; x++) { + lookup_block(ip, x, mp, create, new, dblock); + if (!*dblock) + return 0; + + ret = gfs2_meta_indirect_buffer(ip, x+1, *dblock, *new, &mp->mp_bh[x+1]); + if (ret) + return ret; + } + + return lookup_block(ip, end_of_metadata, mp, create, new, dblock); +} + +static void release_metapath(struct metapath *mp) +{ + int i; + + for (i = 0; i < GFS2_MAX_META_HEIGHT; i++) + if (mp->mp_bh[i]) + brelse(mp->mp_bh[i]); +} + +/** + * gfs2_extent_length - Returns length of an extent of blocks + * @start: Start of the buffer + * @len: Length of the buffer in bytes + * @ptr: Current position in the buffer + * @limit: Max extent length to return (0 = unlimited) + * @eob: Set to 1 if we hit "end of block" + * + * If the first block is zero (unallocated) it will return the number of + * unallocated blocks in the extent, otherwise it will return the number + * of contiguous blocks in the extent. + * + * Returns: The length of the extent (minimum of one block) + */ + +static inline unsigned int gfs2_extent_length(void *start, unsigned int len, __be64 *ptr, unsigned limit, int *eob) +{ + const __be64 *end = (start + len); + const __be64 *first = ptr; + u64 d = be64_to_cpu(*ptr); + + *eob = 0; + do { + ptr++; + if (ptr >= end) + break; + if (limit && --limit == 0) + break; + if (d) + d++; + } while(be64_to_cpu(*ptr) == d); + if (ptr >= end) + *eob = 1; + return (ptr - first); +} + static inline void bmap_lock(struct inode *inode, int create) { struct gfs2_inode *ip = GFS2_I(inode); @@ -457,11 +480,7 @@ int gfs2_block_map(struct inode *inode, sector_t lblock, { struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); - struct buffer_head *bh; - unsigned int bsize; - unsigned int height; - unsigned int end_of_metadata; - unsigned int x; + unsigned int bsize = sdp->sd_sb.sb_bsize; int error = 0; int new = 0; u64 dblock = 0; @@ -469,85 +488,76 @@ int gfs2_block_map(struct inode *inode, sector_t lblock, unsigned int maxlen = bh_map->b_size >> inode->i_blkbits; struct metapath mp; u64 size; - struct buffer_head *dibh = NULL; - + const u64 *arr = sdp->sd_heightsize; BUG_ON(maxlen == 0); if (gfs2_assert_warn(sdp, !gfs2_is_stuffed(ip))) return 0; + memset(mp.mp_bh, 0, sizeof(mp.mp_bh)); bmap_lock(inode, create); clear_buffer_mapped(bh_map); clear_buffer_new(bh_map); clear_buffer_boundary(bh_map); - bsize = gfs2_is_dir(ip) ? sdp->sd_jbsize : sdp->sd_sb.sb_bsize; - size = (lblock + 1) * bsize; - - if (size > ip->i_di.di_size) { - height = calc_tree_height(ip, size); - if (ip->i_di.di_height < height) { - if (!create) - goto out_ok; - - error = build_height(inode, height); - if (error) - goto out_fail; - } + if (gfs2_is_dir(ip)) { + bsize = sdp->sd_jbsize; + arr = sdp->sd_jheightsize; } + size = (lblock + 1) * bsize; - find_metapath(ip, lblock, &mp); - end_of_metadata = ip->i_di.di_height - 1; - error = gfs2_meta_inode_buffer(ip, &bh); + error = gfs2_meta_inode_buffer(ip, &mp.mp_bh[0]); if (error) goto out_fail; - dibh = bh; - get_bh(dibh); - for (x = 0; x < end_of_metadata; x++) { - lookup_block(ip, bh, x, &mp, create, &new, &dblock); - brelse(bh); - if (!dblock) + if (size > arr[ip->i_height]) { + u8 height = ip->i_height; + if (!create) goto out_ok; - - error = gfs2_meta_indirect_buffer(ip, x+1, dblock, new, &bh); + while (size > arr[height]) + height++; + error = build_height(inode, &mp, height); if (error) goto out_fail; } - boundary = lookup_block(ip, bh, end_of_metadata, &mp, create, &new, &dblock); - if (dblock) { + find_metapath(ip, lblock, &mp); + error = lookup_metapath(inode, &mp, create, &new, &dblock); + if (error < 0) + goto out_fail; + boundary = error; + + if (new) { map_bh(bh_map, inode->i_sb, dblock); if (boundary) set_buffer_boundary(bh_map); - if (new) { - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_dinode_out(ip, dibh->b_data); - set_buffer_new(bh_map); - goto out_brelse; - } - while(--maxlen && !buffer_boundary(bh_map)) { - u64 eblock; + gfs2_trans_add_bh(ip->i_gl, mp.mp_bh[0], 1); + gfs2_dinode_out(ip, mp.mp_bh[0]->b_data); + set_buffer_new(bh_map); + goto out_ok; + } - mp.mp_list[end_of_metadata]++; - boundary = lookup_block(ip, bh, end_of_metadata, &mp, 0, &new, &eblock); - if (eblock != ++dblock) - break; - bh_map->b_size += (1 << inode->i_blkbits); - if (boundary) - set_buffer_boundary(bh_map); - } + if (dblock) { + unsigned int len; + struct buffer_head *bh = mp.mp_bh[ip->i_height - 1]; + __be64 *ptr = metapointer(&boundary, ip->i_height - 1, &mp); + map_bh(bh_map, inode->i_sb, dblock); + len = gfs2_extent_length(bh->b_data, bh->b_size, ptr, maxlen, + &boundary); + bh_map->b_size = (len << inode->i_blkbits); + if (boundary) + set_buffer_boundary(bh_map); } -out_brelse: - brelse(bh); out_ok: error = 0; out_fail: - if (dibh) - brelse(dibh); + release_metapath(&mp); bmap_unlock(inode, create); return error; } +/* + * Deprecated: do not use in new code + */ int gfs2_extent_map(struct inode *inode, u64 lblock, int *new, u64 *dblock, unsigned *extlen) { struct buffer_head bh = { .b_state = 0, .b_blocknr = 0 }; @@ -621,7 +631,7 @@ static int recursive_scan(struct gfs2_inode *ip, struct buffer_head *dibh, if (error) goto out; - if (height < ip->i_di.di_height - 1) + if (height < ip->i_height - 1) for (; top < bottom; top++, first = 0) { if (!*top) continue; @@ -679,7 +689,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, sm->sm_first = 0; } - metadata = (height != ip->i_di.di_height - 1); + metadata = (height != ip->i_height - 1); if (metadata) revokes = (height) ? sdp->sd_inptrs : sdp->sd_diptrs; @@ -713,7 +723,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, else goto out; /* Nothing to do */ - gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE, 0); + gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE); for (x = 0; x < rlist.rl_rgrps; x++) { struct gfs2_rgrpd *rgd; @@ -760,10 +770,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, } *p = 0; - if (!ip->i_di.di_blocks) - gfs2_consist_inode(ip); - ip->i_di.di_blocks--; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, -1); } if (bstart) { if (metadata) @@ -804,7 +811,6 @@ static int do_grow(struct gfs2_inode *ip, u64 size) struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_alloc *al; struct buffer_head *dibh; - unsigned int h; int error; al = gfs2_alloc_get(ip); @@ -830,20 +836,24 @@ static int do_grow(struct gfs2_inode *ip, u64 size) goto out_ipres; if (size > sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)) { + const u64 *arr = sdp->sd_heightsize; if (gfs2_is_stuffed(ip)) { error = gfs2_unstuff_dinode(ip, NULL); if (error) goto out_end_trans; } - h = calc_tree_height(ip, size); - if (ip->i_di.di_height < h) { - down_write(&ip->i_rw_mutex); - error = build_height(&ip->i_inode, h); - up_write(&ip->i_rw_mutex); - if (error) - goto out_end_trans; + down_write(&ip->i_rw_mutex); + if (size > arr[ip->i_height]) { + struct metapath mp; + u8 height = ip->i_height; + while(size > arr[height]) + height++; + error = build_height(&ip->i_inode, &mp, height); } + up_write(&ip->i_rw_mutex); + if (error) + goto out_end_trans; } ip->i_di.di_size = size; @@ -986,7 +996,7 @@ out: static int trunc_dealloc(struct gfs2_inode *ip, u64 size) { - unsigned int height = ip->i_di.di_height; + unsigned int height = ip->i_height; u64 lblock; struct metapath mp; int error; @@ -1037,10 +1047,8 @@ static int trunc_end(struct gfs2_inode *ip) goto out; if (!ip->i_di.di_size) { - ip->i_di.di_height = 0; - ip->i_di.di_goal_meta = - ip->i_di.di_goal_data = - ip->i_no_addr; + ip->i_height = 0; + ip->i_goal = ip->i_no_addr; gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); } ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; @@ -1197,10 +1205,9 @@ int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, unsigned int len, int *alloc_required) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - u64 lblock, lblock_stop, dblock; - u32 extlen; - int new = 0; - int error = 0; + struct buffer_head bh; + unsigned int shift; + u64 lblock, lblock_stop, size; *alloc_required = 0; @@ -1214,6 +1221,8 @@ int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, return 0; } + *alloc_required = 1; + shift = sdp->sd_sb.sb_bsize_shift; if (gfs2_is_dir(ip)) { unsigned int bsize = sdp->sd_jbsize; lblock = offset; @@ -1221,27 +1230,25 @@ int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, lblock_stop = offset + len + bsize - 1; do_div(lblock_stop, bsize); } else { - unsigned int shift = sdp->sd_sb.sb_bsize_shift; u64 end_of_file = (ip->i_di.di_size + sdp->sd_sb.sb_bsize - 1) >> shift; lblock = offset >> shift; lblock_stop = (offset + len + sdp->sd_sb.sb_bsize - 1) >> shift; - if (lblock_stop > end_of_file) { - *alloc_required = 1; + if (lblock_stop > end_of_file) return 0; - } } - for (; lblock < lblock_stop; lblock += extlen) { - error = gfs2_extent_map(&ip->i_inode, lblock, &new, &dblock, &extlen); - if (error) - return error; - - if (!dblock) { - *alloc_required = 1; + size = (lblock_stop - lblock) << shift; + do { + bh.b_state = 0; + bh.b_size = size; + gfs2_block_map(&ip->i_inode, lblock, &bh, 0); + if (!buffer_mapped(&bh)) return 0; - } - } + size -= bh.b_size; + lblock += (bh.b_size >> ip->i_inode.i_blkbits); + } while(size > 0); + *alloc_required = 0; return 0; } diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index c347095..34dc8df 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -757,7 +757,7 @@ static struct gfs2_dirent *gfs2_dirent_search(struct inode *inode, if (ip->i_di.di_flags & GFS2_DIF_EXHASH) { struct gfs2_leaf *leaf; - unsigned hsize = 1 << ip->i_di.di_depth; + unsigned hsize = 1 << ip->i_depth; unsigned index; u64 ln; if (hsize * sizeof(u64) != ip->i_di.di_size) { @@ -765,7 +765,7 @@ static struct gfs2_dirent *gfs2_dirent_search(struct inode *inode, return ERR_PTR(-EIO); } - index = name->hash >> (32 - ip->i_di.di_depth); + index = name->hash >> (32 - ip->i_depth); error = get_first_leaf(ip, index, &bh); if (error) return ERR_PTR(error); @@ -803,14 +803,15 @@ got_dent: static struct gfs2_leaf *new_leaf(struct inode *inode, struct buffer_head **pbh, u16 depth) { struct gfs2_inode *ip = GFS2_I(inode); - u64 bn = gfs2_alloc_meta(ip); + unsigned int n = 1; + u64 bn = gfs2_alloc_block(ip, &n); struct buffer_head *bh = gfs2_meta_new(ip->i_gl, bn); struct gfs2_leaf *leaf; struct gfs2_dirent *dent; struct qstr name = { .name = "", .len = 0, .hash = 0 }; if (!bh) return NULL; - + gfs2_trans_add_unrevoke(GFS2_SB(inode), bn, 1); gfs2_trans_add_bh(ip->i_gl, bh, 1); gfs2_metatype_set(bh, GFS2_METATYPE_LF, GFS2_FORMAT_LF); leaf = (struct gfs2_leaf *)bh->b_data; @@ -905,12 +906,11 @@ static int dir_make_exhash(struct inode *inode) *lp = cpu_to_be64(bn); dip->i_di.di_size = sdp->sd_sb.sb_bsize / 2; - dip->i_di.di_blocks++; - gfs2_set_inode_blocks(&dip->i_inode); + gfs2_add_inode_blocks(&dip->i_inode, 1); dip->i_di.di_flags |= GFS2_DIF_EXHASH; for (x = sdp->sd_hash_ptrs, y = -1; x; x >>= 1, y++) ; - dip->i_di.di_depth = y; + dip->i_depth = y; gfs2_dinode_out(dip, dibh->b_data); @@ -941,7 +941,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) int x, moved = 0; int error; - index = name->hash >> (32 - dip->i_di.di_depth); + index = name->hash >> (32 - dip->i_depth); error = get_leaf_nr(dip, index, &leaf_no); if (error) return error; @@ -952,7 +952,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) return error; oleaf = (struct gfs2_leaf *)obh->b_data; - if (dip->i_di.di_depth == be16_to_cpu(oleaf->lf_depth)) { + if (dip->i_depth == be16_to_cpu(oleaf->lf_depth)) { brelse(obh); return 1; /* can't split */ } @@ -967,10 +967,10 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) bn = nbh->b_blocknr; /* Compute the start and len of leaf pointers in the hash table. */ - len = 1 << (dip->i_di.di_depth - be16_to_cpu(oleaf->lf_depth)); + len = 1 << (dip->i_depth - be16_to_cpu(oleaf->lf_depth)); half_len = len >> 1; if (!half_len) { - printk(KERN_WARNING "di_depth %u lf_depth %u index %u\n", dip->i_di.di_depth, be16_to_cpu(oleaf->lf_depth), index); + printk(KERN_WARNING "i_depth %u lf_depth %u index %u\n", dip->i_depth, be16_to_cpu(oleaf->lf_depth), index); gfs2_consist_inode(dip); error = -EIO; goto fail_brelse; @@ -997,7 +997,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) kfree(lp); /* Compute the divider */ - divider = (start + half_len) << (32 - dip->i_di.di_depth); + divider = (start + half_len) << (32 - dip->i_depth); /* Copy the entries */ dirent_first(dip, obh, &dent); @@ -1021,13 +1021,13 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) new->de_inum = dent->de_inum; /* No endian worries */ new->de_type = dent->de_type; /* No endian worries */ - nleaf->lf_entries = cpu_to_be16(be16_to_cpu(nleaf->lf_entries)+1); + be16_add_cpu(&nleaf->lf_entries, 1); dirent_del(dip, obh, prev, dent); if (!oleaf->lf_entries) gfs2_consist_inode(dip); - oleaf->lf_entries = cpu_to_be16(be16_to_cpu(oleaf->lf_entries)-1); + be16_add_cpu(&oleaf->lf_entries, -1); if (!prev) prev = dent; @@ -1044,8 +1044,7 @@ static int dir_split_leaf(struct inode *inode, const struct qstr *name) error = gfs2_meta_inode_buffer(dip, &dibh); if (!gfs2_assert_withdraw(GFS2_SB(&dip->i_inode), !error)) { gfs2_trans_add_bh(dip->i_gl, dibh, 1); - dip->i_di.di_blocks++; - gfs2_set_inode_blocks(&dip->i_inode); + gfs2_add_inode_blocks(&dip->i_inode, 1); gfs2_dinode_out(dip, dibh->b_data); brelse(dibh); } @@ -1082,7 +1081,7 @@ static int dir_double_exhash(struct gfs2_inode *dip) int x; int error = 0; - hsize = 1 << dip->i_di.di_depth; + hsize = 1 << dip->i_depth; if (hsize * sizeof(u64) != dip->i_di.di_size) { gfs2_consist_inode(dip); return -EIO; @@ -1125,7 +1124,7 @@ static int dir_double_exhash(struct gfs2_inode *dip) error = gfs2_meta_inode_buffer(dip, &dibh); if (!gfs2_assert_withdraw(sdp, !error)) { - dip->i_di.di_depth++; + dip->i_depth++; gfs2_dinode_out(dip, dibh->b_data); brelse(dibh); } @@ -1370,14 +1369,14 @@ static int dir_e_read(struct inode *inode, u64 *offset, void *opaque, int error = 0; unsigned depth = 0; - hsize = 1 << dip->i_di.di_depth; + hsize = 1 << dip->i_depth; if (hsize * sizeof(u64) != dip->i_di.di_size) { gfs2_consist_inode(dip); return -EIO; } hash = gfs2_dir_offset2hash(*offset); - index = hash >> (32 - dip->i_di.di_depth); + index = hash >> (32 - dip->i_depth); lp = kmalloc(sdp->sd_hash_bsize, GFP_KERNEL); if (!lp) @@ -1405,7 +1404,7 @@ static int dir_e_read(struct inode *inode, u64 *offset, void *opaque, if (error) break; - len = 1 << (dip->i_di.di_depth - depth); + len = 1 << (dip->i_depth - depth); index = (index & ~(len - 1)) + len; } @@ -1549,7 +1548,7 @@ static int dir_new_leaf(struct inode *inode, const struct qstr *name) u32 index; u64 bn; - index = name->hash >> (32 - ip->i_di.di_depth); + index = name->hash >> (32 - ip->i_depth); error = get_first_leaf(ip, index, &obh); if (error) return error; @@ -1579,8 +1578,7 @@ static int dir_new_leaf(struct inode *inode, const struct qstr *name) if (error) return error; gfs2_trans_add_bh(ip->i_gl, bh, 1); - ip->i_di.di_blocks++; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, 1); gfs2_dinode_out(ip, bh->b_data); brelse(bh); return 0; @@ -1616,7 +1614,7 @@ int gfs2_dir_add(struct inode *inode, const struct qstr *name, dent->de_type = cpu_to_be16(type); if (ip->i_di.di_flags & GFS2_DIF_EXHASH) { leaf = (struct gfs2_leaf *)bh->b_data; - leaf->lf_entries = cpu_to_be16(be16_to_cpu(leaf->lf_entries) + 1); + be16_add_cpu(&leaf->lf_entries, 1); } brelse(bh); error = gfs2_meta_inode_buffer(ip, &bh); @@ -1641,7 +1639,7 @@ int gfs2_dir_add(struct inode *inode, const struct qstr *name, continue; if (error < 0) break; - if (ip->i_di.di_depth < GFS2_DIR_MAX_DEPTH) { + if (ip->i_depth < GFS2_DIR_MAX_DEPTH) { error = dir_double_exhash(ip); if (error) break; @@ -1785,7 +1783,7 @@ static int foreach_leaf(struct gfs2_inode *dip, leaf_call_t lc, void *data) u64 leaf_no; int error = 0; - hsize = 1 << dip->i_di.di_depth; + hsize = 1 << dip->i_depth; if (hsize * sizeof(u64) != dip->i_di.di_size) { gfs2_consist_inode(dip); return -EIO; @@ -1817,7 +1815,7 @@ static int foreach_leaf(struct gfs2_inode *dip, leaf_call_t lc, void *data) if (error) goto out; leaf = (struct gfs2_leaf *)bh->b_data; - len = 1 << (dip->i_di.di_depth - be16_to_cpu(leaf->lf_depth)); + len = 1 << (dip->i_depth - be16_to_cpu(leaf->lf_depth)); brelse(bh); error = lc(dip, index, len, leaf_no, data); @@ -1894,7 +1892,7 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, l_blocks++; } - gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE, 0); + gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE); for (x = 0; x < rlist.rl_rgrps; x++) { struct gfs2_rgrpd *rgd; @@ -1921,11 +1919,7 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, brelse(bh); gfs2_free_meta(dip, blk, 1); - - if (!dip->i_di.di_blocks) - gfs2_consist_inode(dip); - dip->i_di.di_blocks--; - gfs2_set_inode_blocks(&dip->i_inode); + gfs2_add_inode_blocks(&dip->i_inode, -1); } error = gfs2_dir_write_data(dip, ht, index * sizeof(u64), size); diff --git a/fs/gfs2/eattr.c b/fs/gfs2/eattr.c index bee9970..76ead1a 100644 --- a/fs/gfs2/eattr.c +++ b/fs/gfs2/eattr.c @@ -277,10 +277,7 @@ static int ea_dealloc_unstuffed(struct gfs2_inode *ip, struct buffer_head *bh, } *dataptrs = 0; - if (!ip->i_di.di_blocks) - gfs2_consist_inode(ip); - ip->i_di.di_blocks--; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, -1); } if (bstart) gfs2_free_meta(ip, bstart, blen); @@ -582,10 +579,11 @@ static int ea_alloc_blk(struct gfs2_inode *ip, struct buffer_head **bhp) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_ea_header *ea; + unsigned int n = 1; u64 block; - block = gfs2_alloc_meta(ip); - + block = gfs2_alloc_block(ip, &n); + gfs2_trans_add_unrevoke(sdp, block, 1); *bhp = gfs2_meta_new(ip->i_gl, block); gfs2_trans_add_bh(ip->i_gl, *bhp, 1); gfs2_metatype_set(*bhp, GFS2_METATYPE_EA, GFS2_FORMAT_EA); @@ -597,8 +595,7 @@ static int ea_alloc_blk(struct gfs2_inode *ip, struct buffer_head **bhp) ea->ea_flags = GFS2_EAFLAG_LAST; ea->ea_num_ptrs = 0; - ip->i_di.di_blocks++; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, 1); return 0; } @@ -642,15 +639,15 @@ static int ea_write(struct gfs2_inode *ip, struct gfs2_ea_header *ea, struct buffer_head *bh; u64 block; int mh_size = sizeof(struct gfs2_meta_header); + unsigned int n = 1; - block = gfs2_alloc_meta(ip); - + block = gfs2_alloc_block(ip, &n); + gfs2_trans_add_unrevoke(sdp, block, 1); bh = gfs2_meta_new(ip->i_gl, block); gfs2_trans_add_bh(ip->i_gl, bh, 1); gfs2_metatype_set(bh, GFS2_METATYPE_ED, GFS2_FORMAT_ED); - ip->i_di.di_blocks++; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, 1); copy = data_len > sdp->sd_jbsize ? sdp->sd_jbsize : data_len; @@ -966,9 +963,9 @@ static int ea_set_block(struct gfs2_inode *ip, struct gfs2_ea_request *er, gfs2_trans_add_bh(ip->i_gl, indbh, 1); } else { u64 blk; - - blk = gfs2_alloc_meta(ip); - + unsigned int n = 1; + blk = gfs2_alloc_block(ip, &n); + gfs2_trans_add_unrevoke(sdp, blk, 1); indbh = gfs2_meta_new(ip->i_gl, blk); gfs2_trans_add_bh(ip->i_gl, indbh, 1); gfs2_metatype_set(indbh, GFS2_METATYPE_IN, GFS2_FORMAT_IN); @@ -978,8 +975,7 @@ static int ea_set_block(struct gfs2_inode *ip, struct gfs2_ea_request *er, *eablk = cpu_to_be64(ip->i_di.di_eattr); ip->i_di.di_eattr = blk; ip->i_di.di_flags |= GFS2_DIF_EA_INDIRECT; - ip->i_di.di_blocks++; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, 1); eablk++; } @@ -1347,7 +1343,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip) else goto out; - gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE, 0); + gfs2_rlist_alloc(&rlist, LM_ST_EXCLUSIVE); for (x = 0; x < rlist.rl_rgrps; x++) { struct gfs2_rgrpd *rgd; @@ -1387,10 +1383,7 @@ static int ea_dealloc_indirect(struct gfs2_inode *ip) } *eablk = 0; - if (!ip->i_di.di_blocks) - gfs2_consist_inode(ip); - ip->i_di.di_blocks--; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, -1); } if (bstart) gfs2_free_meta(ip, bstart, blen); @@ -1442,10 +1435,7 @@ static int ea_dealloc_block(struct gfs2_inode *ip) gfs2_free_meta(ip, ip->i_di.di_eattr, 1); ip->i_di.di_eattr = 0; - if (!ip->i_di.di_blocks) - gfs2_consist_inode(ip); - ip->i_di.di_blocks--; - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_add_inode_blocks(&ip->i_inode, -1); error = gfs2_meta_inode_buffer(ip, &dibh); if (!error) { diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 7175a4d..63981e2 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -35,7 +35,6 @@ #include "glock.h" #include "glops.h" #include "inode.h" -#include "lm.h" #include "lops.h" #include "meta_io.h" #include "quota.h" @@ -183,7 +182,8 @@ static void glock_free(struct gfs2_glock *gl) struct gfs2_sbd *sdp = gl->gl_sbd; struct inode *aspace = gl->gl_aspace; - gfs2_lm_put_lock(sdp, gl->gl_lock); + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + sdp->sd_lockstruct.ls_ops->lm_put_lock(gl->gl_lock); if (aspace) gfs2_aspace_put(aspace); @@ -197,7 +197,7 @@ static void glock_free(struct gfs2_glock *gl) * */ -void gfs2_glock_hold(struct gfs2_glock *gl) +static void gfs2_glock_hold(struct gfs2_glock *gl) { atomic_inc(&gl->gl_ref); } @@ -293,6 +293,16 @@ static void glock_work_func(struct work_struct *work) gfs2_glock_put(gl); } +static int gfs2_lm_get_lock(struct gfs2_sbd *sdp, struct lm_lockname *name, + void **lockp) +{ + int error = -EIO; + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + error = sdp->sd_lockstruct.ls_ops->lm_get_lock( + sdp->sd_lockstruct.ls_lockspace, name, lockp); + return error; +} + /** * gfs2_glock_get() - Get a glock, or create one if one doesn't exist * @sdp: The GFS2 superblock @@ -338,8 +348,6 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, gl->gl_ip = 0; gl->gl_ops = glops; gl->gl_req_gh = NULL; - gl->gl_req_bh = NULL; - gl->gl_vn = 0; gl->gl_stamp = jiffies; gl->gl_tchange = jiffies; gl->gl_object = NULL; @@ -595,11 +603,12 @@ static void run_queue(struct gfs2_glock *gl) blocked = rq_mutex(gh); } else if (test_bit(GLF_DEMOTE, &gl->gl_flags)) { blocked = rq_demote(gl); - if (gl->gl_waiters2 && !blocked) { + if (test_bit(GLF_WAITERS2, &gl->gl_flags) && + !blocked) { set_bit(GLF_DEMOTE, &gl->gl_flags); gl->gl_demote_state = LM_ST_UNLOCKED; } - gl->gl_waiters2 = 0; + clear_bit(GLF_WAITERS2, &gl->gl_flags); } else if (!list_empty(&gl->gl_waiters3)) { gh = list_entry(gl->gl_waiters3.next, struct gfs2_holder, gh_list); @@ -710,7 +719,7 @@ static void handle_callback(struct gfs2_glock *gl, unsigned int state, } else if (gl->gl_demote_state != LM_ST_UNLOCKED && gl->gl_demote_state != state) { if (test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags)) - gl->gl_waiters2 = 1; + set_bit(GLF_WAITERS2, &gl->gl_flags); else gl->gl_demote_state = LM_ST_UNLOCKED; } @@ -743,6 +752,37 @@ static void state_change(struct gfs2_glock *gl, unsigned int new_state) } /** + * drop_bh - Called after a lock module unlock completes + * @gl: the glock + * @ret: the return status + * + * Doesn't wake up the process waiting on the struct gfs2_holder (if any) + * Doesn't drop the reference on the glock the top half took out + * + */ + +static void drop_bh(struct gfs2_glock *gl, unsigned int ret) +{ + struct gfs2_sbd *sdp = gl->gl_sbd; + const struct gfs2_glock_operations *glops = gl->gl_ops; + + gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags)); + gfs2_assert_warn(sdp, list_empty(&gl->gl_holders)); + gfs2_assert_warn(sdp, !ret); + + state_change(gl, LM_ST_UNLOCKED); + + if (glops->go_inval) + glops->go_inval(gl, DIO_METADATA); + + spin_lock(&gl->gl_spin); + gfs2_demote_wake(gl); + clear_bit(GLF_LOCK, &gl->gl_flags); + spin_unlock(&gl->gl_spin); + gfs2_glock_put(gl); +} + +/** * xmote_bh - Called after the lock module is done acquiring a lock * @gl: The glock in question * @ret: the int returned from the lock module @@ -757,6 +797,11 @@ static void xmote_bh(struct gfs2_glock *gl, unsigned int ret) int prev_state = gl->gl_state; int op_done = 1; + if (!gh && (ret & LM_OUT_ST_MASK) == LM_ST_UNLOCKED) { + drop_bh(gl, ret); + return; + } + gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags)); gfs2_assert_warn(sdp, list_empty(&gl->gl_holders)); gfs2_assert_warn(sdp, !(ret & LM_OUT_ASYNC)); @@ -782,7 +827,6 @@ static void xmote_bh(struct gfs2_glock *gl, unsigned int ret) } else { spin_lock(&gl->gl_spin); if (gl->gl_state != gl->gl_demote_state) { - gl->gl_req_bh = NULL; spin_unlock(&gl->gl_spin); gfs2_glock_drop_th(gl); gfs2_glock_put(gl); @@ -824,7 +868,6 @@ out: if (op_done) { spin_lock(&gl->gl_spin); gl->gl_req_gh = NULL; - gl->gl_req_bh = NULL; clear_bit(GLF_LOCK, &gl->gl_flags); spin_unlock(&gl->gl_spin); } @@ -835,6 +878,17 @@ out: gfs2_holder_wake(gh); } +static unsigned int gfs2_lm_lock(struct gfs2_sbd *sdp, void *lock, + unsigned int cur_state, unsigned int req_state, + unsigned int flags) +{ + int ret = 0; + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + ret = sdp->sd_lockstruct.ls_ops->lm_lock(lock, cur_state, + req_state, flags); + return ret; +} + /** * gfs2_glock_xmote_th - Call into the lock module to acquire or change a glock * @gl: The glock in question @@ -863,7 +917,6 @@ static void gfs2_glock_xmote_th(struct gfs2_glock *gl, struct gfs2_holder *gh) gfs2_assert_warn(sdp, state != gl->gl_state); gfs2_glock_hold(gl); - gl->gl_req_bh = xmote_bh; lck_ret = gfs2_lm_lock(sdp, gl->gl_lock, gl->gl_state, state, lck_flags); @@ -876,49 +929,13 @@ static void gfs2_glock_xmote_th(struct gfs2_glock *gl, struct gfs2_holder *gh) xmote_bh(gl, lck_ret); } -/** - * drop_bh - Called after a lock module unlock completes - * @gl: the glock - * @ret: the return status - * - * Doesn't wake up the process waiting on the struct gfs2_holder (if any) - * Doesn't drop the reference on the glock the top half took out - * - */ - -static void drop_bh(struct gfs2_glock *gl, unsigned int ret) +static unsigned int gfs2_lm_unlock(struct gfs2_sbd *sdp, void *lock, + unsigned int cur_state) { - struct gfs2_sbd *sdp = gl->gl_sbd; - const struct gfs2_glock_operations *glops = gl->gl_ops; - struct gfs2_holder *gh = gl->gl_req_gh; - - gfs2_assert_warn(sdp, test_bit(GLF_LOCK, &gl->gl_flags)); - gfs2_assert_warn(sdp, list_empty(&gl->gl_holders)); - gfs2_assert_warn(sdp, !ret); - - state_change(gl, LM_ST_UNLOCKED); - - if (glops->go_inval) - glops->go_inval(gl, DIO_METADATA); - - if (gh) { - spin_lock(&gl->gl_spin); - list_del_init(&gh->gh_list); - gh->gh_error = 0; - spin_unlock(&gl->gl_spin); - } - - spin_lock(&gl->gl_spin); - gfs2_demote_wake(gl); - gl->gl_req_gh = NULL; - gl->gl_req_bh = NULL; - clear_bit(GLF_LOCK, &gl->gl_flags); - spin_unlock(&gl->gl_spin); - - gfs2_glock_put(gl); - - if (gh) - gfs2_holder_wake(gh); + int ret = 0; + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + ret = sdp->sd_lockstruct.ls_ops->lm_unlock(lock, cur_state); + return ret; } /** @@ -941,7 +958,6 @@ static void gfs2_glock_drop_th(struct gfs2_glock *gl) gfs2_assert_warn(sdp, gl->gl_state != LM_ST_UNLOCKED); gfs2_glock_hold(gl); - gl->gl_req_bh = drop_bh; ret = gfs2_lm_unlock(sdp, gl->gl_lock, gl->gl_state); @@ -964,16 +980,17 @@ static void gfs2_glock_drop_th(struct gfs2_glock *gl) static void do_cancels(struct gfs2_holder *gh) { struct gfs2_glock *gl = gh->gh_gl; + struct gfs2_sbd *sdp = gl->gl_sbd; spin_lock(&gl->gl_spin); while (gl->gl_req_gh != gh && !test_bit(HIF_HOLDER, &gh->gh_iflags) && !list_empty(&gh->gh_list)) { - if (gl->gl_req_bh && !(gl->gl_req_gh && - (gl->gl_req_gh->gh_flags & GL_NOCANCEL))) { + if (!(gl->gl_req_gh && (gl->gl_req_gh->gh_flags & GL_NOCANCEL))) { spin_unlock(&gl->gl_spin); - gfs2_lm_cancel(gl->gl_sbd, gl->gl_lock); + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + sdp->sd_lockstruct.ls_ops->lm_cancel(gl->gl_lock); msleep(100); spin_lock(&gl->gl_spin); } else { @@ -1041,7 +1058,6 @@ static int glock_wait_internal(struct gfs2_holder *gh) spin_lock(&gl->gl_spin); gl->gl_req_gh = NULL; - gl->gl_req_bh = NULL; clear_bit(GLF_LOCK, &gl->gl_flags); run_queue(gl); spin_unlock(&gl->gl_spin); @@ -1428,6 +1444,14 @@ void gfs2_glock_dq_uninit_m(unsigned int num_gh, struct gfs2_holder *ghs) gfs2_glock_dq_uninit(&ghs[x]); } +static int gfs2_lm_hold_lvb(struct gfs2_sbd *sdp, void *lock, char **lvbp) +{ + int error = -EIO; + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + error = sdp->sd_lockstruct.ls_ops->lm_hold_lvb(lock, lvbp); + return error; +} + /** * gfs2_lvb_hold - attach a LVB from a glock * @gl: The glock in question @@ -1463,12 +1487,15 @@ int gfs2_lvb_hold(struct gfs2_glock *gl) void gfs2_lvb_unhold(struct gfs2_glock *gl) { + struct gfs2_sbd *sdp = gl->gl_sbd; + gfs2_glock_hold(gl); gfs2_glmutex_lock(gl); gfs2_assert(gl->gl_sbd, atomic_read(&gl->gl_lvb_count) > 0); if (atomic_dec_and_test(&gl->gl_lvb_count)) { - gfs2_lm_unhold_lvb(gl->gl_sbd, gl->gl_lock, gl->gl_lvb); + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + sdp->sd_lockstruct.ls_ops->lm_unhold_lvb(gl->gl_lock, gl->gl_lvb); gl->gl_lvb = NULL; gfs2_glock_put(gl); } @@ -1534,8 +1561,7 @@ void gfs2_glock_cb(void *cb_data, unsigned int type, void *data) gl = gfs2_glock_find(sdp, &async->lc_name); if (gfs2_assert_warn(sdp, gl)) return; - if (!gfs2_assert_warn(sdp, gl->gl_req_bh)) - gl->gl_req_bh(gl, async->lc_ret); + xmote_bh(gl, async->lc_ret); if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) gfs2_glock_put(gl); up_read(&gfs2_umount_flush_sem); @@ -1594,10 +1620,10 @@ void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) gfs2_glock_hold(gl); list_add(&gl->gl_reclaim, &sdp->sd_reclaim_list); atomic_inc(&sdp->sd_reclaim_count); - } - spin_unlock(&sdp->sd_reclaim_lock); - - wake_up(&sdp->sd_reclaim_wq); + spin_unlock(&sdp->sd_reclaim_lock); + wake_up(&sdp->sd_reclaim_wq); + } else + spin_unlock(&sdp->sd_reclaim_lock); } /** @@ -1897,7 +1923,6 @@ static int dump_glock(struct glock_iter *gi, struct gfs2_glock *gl) print_dbg(gi, " gl_owner = -1\n"); print_dbg(gi, " gl_ip = %lu\n", gl->gl_ip); print_dbg(gi, " req_gh = %s\n", (gl->gl_req_gh) ? "yes" : "no"); - print_dbg(gi, " req_bh = %s\n", (gl->gl_req_bh) ? "yes" : "no"); print_dbg(gi, " lvb_count = %d\n", atomic_read(&gl->gl_lvb_count)); print_dbg(gi, " object = %s\n", (gl->gl_object) ? "yes" : "no"); print_dbg(gi, " reclaim = %s\n", diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h index 2f9c6d1..ace5770 100644 --- a/fs/gfs2/glock.h +++ b/fs/gfs2/glock.h @@ -79,7 +79,6 @@ static inline int gfs2_glock_is_blocking(struct gfs2_glock *gl) int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, const struct gfs2_glock_operations *glops, int create, struct gfs2_glock **glp); -void gfs2_glock_hold(struct gfs2_glock *gl); int gfs2_glock_put(struct gfs2_glock *gl); void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags, struct gfs2_holder *gh); diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index c663b7a..d31bada 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -126,7 +126,13 @@ static void meta_go_inval(struct gfs2_glock *gl, int flags) return; gfs2_meta_inval(gl); - gl->gl_vn++; + if (gl->gl_object == GFS2_I(gl->gl_sbd->sd_rindex)) + gl->gl_sbd->sd_rindex_uptodate = 0; + else if (gl->gl_ops == &gfs2_rgrp_glops && gl->gl_object) { + struct gfs2_rgrpd *rgd = (struct gfs2_rgrpd *)gl->gl_object; + + rgd->rd_flags &= ~GFS2_RDF_UPTODATE; + } } /** diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index 525dcae..4ba2ea6 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -44,7 +44,6 @@ struct gfs2_log_header_host { struct gfs2_log_operations { void (*lo_add) (struct gfs2_sbd *sdp, struct gfs2_log_element *le); - void (*lo_incore_commit) (struct gfs2_sbd *sdp, struct gfs2_trans *tr); void (*lo_before_commit) (struct gfs2_sbd *sdp); void (*lo_after_commit) (struct gfs2_sbd *sdp, struct gfs2_ail *ai); void (*lo_before_scan) (struct gfs2_jdesc *jd, @@ -70,7 +69,6 @@ struct gfs2_bitmap { }; struct gfs2_rgrp_host { - u32 rg_flags; u32 rg_free; u32 rg_dinodes; u64 rg_igeneration; @@ -87,17 +85,17 @@ struct gfs2_rgrpd { u32 rd_data; /* num of data blocks in rgrp */ u32 rd_bitbytes; /* number of bytes in data bitmaps */ struct gfs2_rgrp_host rd_rg; - u64 rd_rg_vn; struct gfs2_bitmap *rd_bits; unsigned int rd_bh_count; struct mutex rd_mutex; u32 rd_free_clone; struct gfs2_log_element rd_le; - u32 rd_last_alloc_data; - u32 rd_last_alloc_meta; + u32 rd_last_alloc; struct gfs2_sbd *rd_sbd; - unsigned long rd_flags; -#define GFS2_RDF_CHECK 0x0001 /* Need to check for unlinked inodes */ + unsigned char rd_flags; +#define GFS2_RDF_CHECK 0x01 /* Need to check for unlinked inodes */ +#define GFS2_RDF_NOALLOC 0x02 /* rg prohibits allocation */ +#define GFS2_RDF_UPTODATE 0x04 /* rg is up to date */ }; enum gfs2_state_bits { @@ -168,6 +166,7 @@ enum { GLF_DIRTY = 5, GLF_DEMOTE_IN_PROGRESS = 6, GLF_LFLUSH = 7, + GLF_WAITERS2 = 8, }; struct gfs2_glock { @@ -187,18 +186,15 @@ struct gfs2_glock { struct list_head gl_holders; struct list_head gl_waiters1; /* HIF_MUTEX */ struct list_head gl_waiters3; /* HIF_PROMOTE */ - int gl_waiters2; /* GIF_DEMOTE */ const struct gfs2_glock_operations *gl_ops; struct gfs2_holder *gl_req_gh; - gfs2_glop_bh_t gl_req_bh; void *gl_lock; char *gl_lvb; atomic_t gl_lvb_count; - u64 gl_vn; unsigned long gl_stamp; unsigned long gl_tchange; void *gl_object; @@ -213,6 +209,8 @@ struct gfs2_glock { struct delayed_work gl_work; }; +#define GFS2_MIN_LVB_SIZE 32 /* Min size of LVB that gfs2 supports */ + struct gfs2_alloc { /* Quota stuff */ @@ -241,14 +239,9 @@ enum { struct gfs2_dinode_host { u64 di_size; /* number of bytes in file */ - u64 di_blocks; /* number of blocks in file */ - u64 di_goal_meta; /* rgrp to alloc from next */ - u64 di_goal_data; /* data block goal */ u64 di_generation; /* generation number for NFS */ u32 di_flags; /* GFS2_DIF_... */ - u16 di_height; /* height of metadata */ /* These only apply to directories */ - u16 di_depth; /* Number of bits in the table */ u32 di_entries; /* The number of entries in the directory */ u64 di_eattr; /* extended attribute block number */ }; @@ -265,9 +258,10 @@ struct gfs2_inode { struct gfs2_holder i_iopen_gh; struct gfs2_holder i_gh; /* for prepare/commit_write only */ struct gfs2_alloc *i_alloc; - u64 i_last_rg_alloc; - + u64 i_goal; /* goal block for allocations */ struct rw_semaphore i_rw_mutex; + u8 i_height; + u8 i_depth; }; /* @@ -490,9 +484,9 @@ struct gfs2_sbd { u32 sd_qc_per_block; u32 sd_max_dirres; /* Max blocks needed to add a directory entry */ u32 sd_max_height; /* Max height of a file's metadata tree */ - u64 sd_heightsize[GFS2_MAX_META_HEIGHT]; + u64 sd_heightsize[GFS2_MAX_META_HEIGHT + 1]; u32 sd_max_jheight; /* Max height of journaled file's meta tree */ - u64 sd_jheightsize[GFS2_MAX_META_HEIGHT]; + u64 sd_jheightsize[GFS2_MAX_META_HEIGHT + 1]; struct gfs2_args sd_args; /* Mount arguments */ struct gfs2_tune sd_tune; /* Filesystem tuning structure */ @@ -533,7 +527,7 @@ struct gfs2_sbd { /* Resource group stuff */ - u64 sd_rindex_vn; + int sd_rindex_uptodate; spinlock_t sd_rindex_spin; struct mutex sd_rindex_mutex; struct list_head sd_rindex_list; @@ -637,9 +631,6 @@ struct gfs2_sbd { /* Counters */ - atomic_t sd_glock_count; - atomic_t sd_glock_held_count; - atomic_t sd_inode_count; atomic_t sd_reclaimed; char sd_fsname[GFS2_FSNAME_LEN]; diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 37725ad..5f50dd5 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -248,12 +248,10 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) { struct gfs2_dinode_host *di = &ip->i_di; const struct gfs2_dinode *str = buf; + u16 height, depth; - if (ip->i_no_addr != be64_to_cpu(str->di_num.no_addr)) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; - } + if (unlikely(ip->i_no_addr != be64_to_cpu(str->di_num.no_addr))) + goto corrupt; ip->i_no_formal_ino = be64_to_cpu(str->di_num.no_formal_ino); ip->i_inode.i_mode = be32_to_cpu(str->di_mode); ip->i_inode.i_rdev = 0; @@ -275,8 +273,7 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) ip->i_inode.i_nlink = be32_to_cpu(str->di_nlink); di->di_size = be64_to_cpu(str->di_size); i_size_write(&ip->i_inode, di->di_size); - di->di_blocks = be64_to_cpu(str->di_blocks); - gfs2_set_inode_blocks(&ip->i_inode); + gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks)); ip->i_inode.i_atime.tv_sec = be64_to_cpu(str->di_atime); ip->i_inode.i_atime.tv_nsec = be32_to_cpu(str->di_atime_nsec); ip->i_inode.i_mtime.tv_sec = be64_to_cpu(str->di_mtime); @@ -284,15 +281,20 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) ip->i_inode.i_ctime.tv_sec = be64_to_cpu(str->di_ctime); ip->i_inode.i_ctime.tv_nsec = be32_to_cpu(str->di_ctime_nsec); - di->di_goal_meta = be64_to_cpu(str->di_goal_meta); - di->di_goal_data = be64_to_cpu(str->di_goal_data); + ip->i_goal = be64_to_cpu(str->di_goal_meta); di->di_generation = be64_to_cpu(str->di_generation); di->di_flags = be32_to_cpu(str->di_flags); gfs2_set_inode_flags(&ip->i_inode); - di->di_height = be16_to_cpu(str->di_height); - - di->di_depth = be16_to_cpu(str->di_depth); + height = be16_to_cpu(str->di_height); + if (unlikely(height > GFS2_MAX_META_HEIGHT)) + goto corrupt; + ip->i_height = (u8)height; + + depth = be16_to_cpu(str->di_depth); + if (unlikely(depth > GFS2_DIR_MAX_DEPTH)) + goto corrupt; + ip->i_depth = (u8)depth; di->di_entries = be32_to_cpu(str->di_entries); di->di_eattr = be64_to_cpu(str->di_eattr); @@ -300,6 +302,10 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) gfs2_set_aops(&ip->i_inode); return 0; +corrupt: + if (gfs2_consist_inode(ip)) + gfs2_dinode_print(ip); + return -EIO; } /** @@ -337,7 +343,7 @@ int gfs2_dinode_dealloc(struct gfs2_inode *ip) struct gfs2_rgrpd *rgd; int error; - if (ip->i_di.di_blocks != 1) { + if (gfs2_get_inode_blocks(&ip->i_inode) != 1) { if (gfs2_consist_inode(ip)) gfs2_dinode_print(ip); return -EIO; @@ -1391,21 +1397,21 @@ void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf) str->di_gid = cpu_to_be32(ip->i_inode.i_gid); str->di_nlink = cpu_to_be32(ip->i_inode.i_nlink); str->di_size = cpu_to_be64(di->di_size); - str->di_blocks = cpu_to_be64(di->di_blocks); + str->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); str->di_atime = cpu_to_be64(ip->i_inode.i_atime.tv_sec); str->di_mtime = cpu_to_be64(ip->i_inode.i_mtime.tv_sec); str->di_ctime = cpu_to_be64(ip->i_inode.i_ctime.tv_sec); - str->di_goal_meta = cpu_to_be64(di->di_goal_meta); - str->di_goal_data = cpu_to_be64(di->di_goal_data); + str->di_goal_meta = cpu_to_be64(ip->i_goal); + str->di_goal_data = cpu_to_be64(ip->i_goal); str->di_generation = cpu_to_be64(di->di_generation); str->di_flags = cpu_to_be32(di->di_flags); - str->di_height = cpu_to_be16(di->di_height); + str->di_height = cpu_to_be16(ip->i_height); str->di_payload_format = cpu_to_be32(S_ISDIR(ip->i_inode.i_mode) && !(ip->i_di.di_flags & GFS2_DIF_EXHASH) ? GFS2_FORMAT_DE : 0); - str->di_depth = cpu_to_be16(di->di_depth); + str->di_depth = cpu_to_be16(ip->i_depth); str->di_entries = cpu_to_be32(di->di_entries); str->di_eattr = cpu_to_be64(di->di_eattr); @@ -1423,15 +1429,13 @@ void gfs2_dinode_print(const struct gfs2_inode *ip) printk(KERN_INFO " no_addr = %llu\n", (unsigned long long)ip->i_no_addr); printk(KERN_INFO " di_size = %llu\n", (unsigned long long)di->di_size); - printk(KERN_INFO " di_blocks = %llu\n", - (unsigned long long)di->di_blocks); - printk(KERN_INFO " di_goal_meta = %llu\n", - (unsigned long long)di->di_goal_meta); - printk(KERN_INFO " di_goal_data = %llu\n", - (unsigned long long)di->di_goal_data); + printk(KERN_INFO " blocks = %llu\n", + (unsigned long long)gfs2_get_inode_blocks(&ip->i_inode)); + printk(KERN_INFO " i_goal = %llu\n", + (unsigned long long)ip->i_goal); printk(KERN_INFO " di_flags = 0x%.8X\n", di->di_flags); - printk(KERN_INFO " di_height = %u\n", di->di_height); - printk(KERN_INFO " di_depth = %u\n", di->di_depth); + printk(KERN_INFO " i_height = %u\n", ip->i_height); + printk(KERN_INFO " i_depth = %u\n", ip->i_depth); printk(KERN_INFO " di_entries = %u\n", di->di_entries); printk(KERN_INFO " di_eattr = %llu\n", (unsigned long long)di->di_eattr); diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h index d446506..580da45 100644 --- a/fs/gfs2/inode.h +++ b/fs/gfs2/inode.h @@ -10,9 +10,11 @@ #ifndef __INODE_DOT_H__ #define __INODE_DOT_H__ +#include "util.h" + static inline int gfs2_is_stuffed(const struct gfs2_inode *ip) { - return !ip->i_di.di_height; + return !ip->i_height; } static inline int gfs2_is_jdata(const struct gfs2_inode *ip) @@ -37,13 +39,25 @@ static inline int gfs2_is_dir(const struct gfs2_inode *ip) return S_ISDIR(ip->i_inode.i_mode); } -static inline void gfs2_set_inode_blocks(struct inode *inode) +static inline void gfs2_set_inode_blocks(struct inode *inode, u64 blocks) +{ + inode->i_blocks = blocks << + (GFS2_SB(inode)->sd_sb.sb_bsize_shift - GFS2_BASIC_BLOCK_SHIFT); +} + +static inline u64 gfs2_get_inode_blocks(const struct inode *inode) { - struct gfs2_inode *ip = GFS2_I(inode); - inode->i_blocks = ip->i_di.di_blocks << + return inode->i_blocks >> (GFS2_SB(inode)->sd_sb.sb_bsize_shift - GFS2_BASIC_BLOCK_SHIFT); } +static inline void gfs2_add_inode_blocks(struct inode *inode, s64 change) +{ + gfs2_assert(GFS2_SB(inode), (change >= 0 || inode->i_blocks > -change)); + change *= (GFS2_SB(inode)->sd_sb.sb_bsize/GFS2_BASIC_BLOCK); + inode->i_blocks += change; +} + static inline int gfs2_check_inum(const struct gfs2_inode *ip, u64 no_addr, u64 no_formal_ino) { diff --git a/fs/gfs2/lm.c b/fs/gfs2/lm.c deleted file mode 100644 index cfcc39b..0000000 --- a/fs/gfs2/lm.c +++ /dev/null @@ -1,210 +0,0 @@ -/* - * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved. - * - * This copyrighted material is made available to anyone wishing to use, - * modify, copy, or redistribute it subject to the terms and conditions - * of the GNU General Public License version 2. - */ - -#include -#include -#include -#include -#include -#include -#include - -#include "gfs2.h" -#include "incore.h" -#include "glock.h" -#include "lm.h" -#include "super.h" -#include "util.h" - -/** - * gfs2_lm_mount - mount a locking protocol - * @sdp: the filesystem - * @args: mount arguements - * @silent: if 1, don't complain if the FS isn't a GFS2 fs - * - * Returns: errno - */ - -int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent) -{ - char *proto = sdp->sd_proto_name; - char *table = sdp->sd_table_name; - int flags = 0; - int error; - - if (sdp->sd_args.ar_spectator) - flags |= LM_MFLAG_SPECTATOR; - - fs_info(sdp, "Trying to join cluster \"%s\", \"%s\"\n", proto, table); - - error = gfs2_mount_lockproto(proto, table, sdp->sd_args.ar_hostdata, - gfs2_glock_cb, sdp, - GFS2_MIN_LVB_SIZE, flags, - &sdp->sd_lockstruct, &sdp->sd_kobj); - if (error) { - fs_info(sdp, "can't mount proto=%s, table=%s, hostdata=%s\n", - proto, table, sdp->sd_args.ar_hostdata); - goto out; - } - - if (gfs2_assert_warn(sdp, sdp->sd_lockstruct.ls_lockspace) || - gfs2_assert_warn(sdp, sdp->sd_lockstruct.ls_ops) || - gfs2_assert_warn(sdp, sdp->sd_lockstruct.ls_lvb_size >= - GFS2_MIN_LVB_SIZE)) { - gfs2_unmount_lockproto(&sdp->sd_lockstruct); - goto out; - } - - if (sdp->sd_args.ar_spectator) - snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s.s", table); - else - snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s.%u", table, - sdp->sd_lockstruct.ls_jid); - - fs_info(sdp, "Joined cluster. Now mounting FS...\n"); - - if ((sdp->sd_lockstruct.ls_flags & LM_LSFLAG_LOCAL) && - !sdp->sd_args.ar_ignore_local_fs) { - sdp->sd_args.ar_localflocks = 1; - sdp->sd_args.ar_localcaching = 1; - } - -out: - return error; -} - -void gfs2_lm_others_may_mount(struct gfs2_sbd *sdp) -{ - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - sdp->sd_lockstruct.ls_ops->lm_others_may_mount( - sdp->sd_lockstruct.ls_lockspace); -} - -void gfs2_lm_unmount(struct gfs2_sbd *sdp) -{ - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - gfs2_unmount_lockproto(&sdp->sd_lockstruct); -} - -int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...) -{ - va_list args; - - if (test_and_set_bit(SDF_SHUTDOWN, &sdp->sd_flags)) - return 0; - - va_start(args, fmt); - vprintk(fmt, args); - va_end(args); - - fs_err(sdp, "about to withdraw this file system\n"); - BUG_ON(sdp->sd_args.ar_debug); - - fs_err(sdp, "telling LM to withdraw\n"); - gfs2_withdraw_lockproto(&sdp->sd_lockstruct); - fs_err(sdp, "withdrawn\n"); - dump_stack(); - - return -1; -} - -int gfs2_lm_get_lock(struct gfs2_sbd *sdp, struct lm_lockname *name, - void **lockp) -{ - int error = -EIO; - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - error = sdp->sd_lockstruct.ls_ops->lm_get_lock( - sdp->sd_lockstruct.ls_lockspace, name, lockp); - return error; -} - -void gfs2_lm_put_lock(struct gfs2_sbd *sdp, void *lock) -{ - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - sdp->sd_lockstruct.ls_ops->lm_put_lock(lock); -} - -unsigned int gfs2_lm_lock(struct gfs2_sbd *sdp, void *lock, - unsigned int cur_state, unsigned int req_state, - unsigned int flags) -{ - int ret = 0; - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - ret = sdp->sd_lockstruct.ls_ops->lm_lock(lock, cur_state, - req_state, flags); - return ret; -} - -unsigned int gfs2_lm_unlock(struct gfs2_sbd *sdp, void *lock, - unsigned int cur_state) -{ - int ret = 0; - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - ret = sdp->sd_lockstruct.ls_ops->lm_unlock(lock, cur_state); - return ret; -} - -void gfs2_lm_cancel(struct gfs2_sbd *sdp, void *lock) -{ - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - sdp->sd_lockstruct.ls_ops->lm_cancel(lock); -} - -int gfs2_lm_hold_lvb(struct gfs2_sbd *sdp, void *lock, char **lvbp) -{ - int error = -EIO; - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - error = sdp->sd_lockstruct.ls_ops->lm_hold_lvb(lock, lvbp); - return error; -} - -void gfs2_lm_unhold_lvb(struct gfs2_sbd *sdp, void *lock, char *lvb) -{ - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - sdp->sd_lockstruct.ls_ops->lm_unhold_lvb(lock, lvb); -} - -int gfs2_lm_plock_get(struct gfs2_sbd *sdp, struct lm_lockname *name, - struct file *file, struct file_lock *fl) -{ - int error = -EIO; - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - error = sdp->sd_lockstruct.ls_ops->lm_plock_get( - sdp->sd_lockstruct.ls_lockspace, name, file, fl); - return error; -} - -int gfs2_lm_plock(struct gfs2_sbd *sdp, struct lm_lockname *name, - struct file *file, int cmd, struct file_lock *fl) -{ - int error = -EIO; - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - error = sdp->sd_lockstruct.ls_ops->lm_plock( - sdp->sd_lockstruct.ls_lockspace, name, file, cmd, fl); - return error; -} - -int gfs2_lm_punlock(struct gfs2_sbd *sdp, struct lm_lockname *name, - struct file *file, struct file_lock *fl) -{ - int error = -EIO; - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - error = sdp->sd_lockstruct.ls_ops->lm_punlock( - sdp->sd_lockstruct.ls_lockspace, name, file, fl); - return error; -} - -void gfs2_lm_recovery_done(struct gfs2_sbd *sdp, unsigned int jid, - unsigned int message) -{ - if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) - sdp->sd_lockstruct.ls_ops->lm_recovery_done( - sdp->sd_lockstruct.ls_lockspace, jid, message); -} - diff --git a/fs/gfs2/lm.h b/fs/gfs2/lm.h deleted file mode 100644 index 21cdc30..0000000 --- a/fs/gfs2/lm.h +++ /dev/null @@ -1,42 +0,0 @@ -/* - * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved. - * - * This copyrighted material is made available to anyone wishing to use, - * modify, copy, or redistribute it subject to the terms and conditions - * of the GNU General Public License version 2. - */ - -#ifndef __LM_DOT_H__ -#define __LM_DOT_H__ - -struct gfs2_sbd; - -#define GFS2_MIN_LVB_SIZE 32 - -int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent); -void gfs2_lm_others_may_mount(struct gfs2_sbd *sdp); -void gfs2_lm_unmount(struct gfs2_sbd *sdp); -int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...) - __attribute__ ((format(printf, 2, 3))); -int gfs2_lm_get_lock(struct gfs2_sbd *sdp, struct lm_lockname *name, - void **lockp); -void gfs2_lm_put_lock(struct gfs2_sbd *sdp, void *lock); -unsigned int gfs2_lm_lock(struct gfs2_sbd *sdp, void *lock, - unsigned int cur_state, unsigned int req_state, - unsigned int flags); -unsigned int gfs2_lm_unlock(struct gfs2_sbd *sdp, void *lock, - unsigned int cur_state); -void gfs2_lm_cancel(struct gfs2_sbd *sdp, void *lock); -int gfs2_lm_hold_lvb(struct gfs2_sbd *sdp, void *lock, char **lvbp); -void gfs2_lm_unhold_lvb(struct gfs2_sbd *sdp, void *lock, char *lvb); -int gfs2_lm_plock_get(struct gfs2_sbd *sdp, struct lm_lockname *name, - struct file *file, struct file_lock *fl); -int gfs2_lm_plock(struct gfs2_sbd *sdp, struct lm_lockname *name, - struct file *file, int cmd, struct file_lock *fl); -int gfs2_lm_punlock(struct gfs2_sbd *sdp, struct lm_lockname *name, - struct file *file, struct file_lock *fl); -void gfs2_lm_recovery_done(struct gfs2_sbd *sdp, unsigned int jid, - unsigned int message); - -#endif /* __LM_DOT_H__ */ diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index 161ab6f..b335304 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -779,6 +779,21 @@ static void log_refund(struct gfs2_sbd *sdp, struct gfs2_trans *tr) gfs2_log_unlock(sdp); } +static void buf_lo_incore_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) +{ + struct list_head *head = &tr->tr_list_buf; + struct gfs2_bufdata *bd; + + gfs2_log_lock(sdp); + while (!list_empty(head)) { + bd = list_entry(head->next, struct gfs2_bufdata, bd_list_tr); + list_del_init(&bd->bd_list_tr); + tr->tr_num_buf--; + } + gfs2_log_unlock(sdp); + gfs2_assert_warn(sdp, !tr->tr_num_buf); +} + /** * gfs2_log_commit - Commit a transaction to the log * @sdp: the filesystem @@ -790,7 +805,7 @@ static void log_refund(struct gfs2_sbd *sdp, struct gfs2_trans *tr) void gfs2_log_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) { log_refund(sdp, tr); - lops_incore_commit(sdp, tr); + buf_lo_incore_commit(sdp, tr); sdp->sd_vfs->s_dirt = 1; up_read(&sdp->sd_log_flush_lock); diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c index fae59d6..4390f6f 100644 --- a/fs/gfs2/lops.c +++ b/fs/gfs2/lops.c @@ -152,21 +152,6 @@ out: unlock_buffer(bd->bd_bh); } -static void buf_lo_incore_commit(struct gfs2_sbd *sdp, struct gfs2_trans *tr) -{ - struct list_head *head = &tr->tr_list_buf; - struct gfs2_bufdata *bd; - - gfs2_log_lock(sdp); - while (!list_empty(head)) { - bd = list_entry(head->next, struct gfs2_bufdata, bd_list_tr); - list_del_init(&bd->bd_list_tr); - tr->tr_num_buf--; - } - gfs2_log_unlock(sdp); - gfs2_assert_warn(sdp, !tr->tr_num_buf); -} - static void buf_lo_before_commit(struct gfs2_sbd *sdp) { struct buffer_head *bh; @@ -419,8 +404,10 @@ static int revoke_lo_scan_elements(struct gfs2_jdesc *jd, unsigned int start, blkno = be64_to_cpu(*(__be64 *)(bh->b_data + offset)); error = gfs2_revoke_add(sdp, blkno, start); - if (error < 0) + if (error < 0) { + brelse(bh); return error; + } else if (error) sdp->sd_found_revokes++; @@ -737,7 +724,6 @@ static void databuf_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai) const struct gfs2_log_operations gfs2_buf_lops = { .lo_add = buf_lo_add, - .lo_incore_commit = buf_lo_incore_commit, .lo_before_commit = buf_lo_before_commit, .lo_after_commit = buf_lo_after_commit, .lo_before_scan = buf_lo_before_scan, @@ -763,7 +749,6 @@ const struct gfs2_log_operations gfs2_rg_lops = { const struct gfs2_log_operations gfs2_databuf_lops = { .lo_add = databuf_lo_add, - .lo_incore_commit = buf_lo_incore_commit, .lo_before_commit = databuf_lo_before_commit, .lo_after_commit = databuf_lo_after_commit, .lo_scan_elements = databuf_lo_scan_elements, diff --git a/fs/gfs2/lops.h b/fs/gfs2/lops.h index 41a00df..3c0b273 100644 --- a/fs/gfs2/lops.h +++ b/fs/gfs2/lops.h @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -57,15 +57,6 @@ static inline void lops_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le) le->le_ops->lo_add(sdp, le); } -static inline void lops_incore_commit(struct gfs2_sbd *sdp, - struct gfs2_trans *tr) -{ - int x; - for (x = 0; gfs2_log_ops[x]; x++) - if (gfs2_log_ops[x]->lo_incore_commit) - gfs2_log_ops[x]->lo_incore_commit(sdp, tr); -} - static inline void lops_before_commit(struct gfs2_sbd *sdp) { int x; diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c index 9c7765c..053e2eb 100644 --- a/fs/gfs2/main.c +++ b/fs/gfs2/main.c @@ -89,6 +89,12 @@ static int __init init_gfs2_fs(void) if (!gfs2_bufdata_cachep) goto fail; + gfs2_rgrpd_cachep = kmem_cache_create("gfs2_rgrpd", + sizeof(struct gfs2_rgrpd), + 0, 0, NULL); + if (!gfs2_rgrpd_cachep) + goto fail; + error = register_filesystem(&gfs2_fs_type); if (error) goto fail; @@ -108,6 +114,9 @@ fail_unregister: fail: gfs2_glock_exit(); + if (gfs2_rgrpd_cachep) + kmem_cache_destroy(gfs2_rgrpd_cachep); + if (gfs2_bufdata_cachep) kmem_cache_destroy(gfs2_bufdata_cachep); @@ -133,6 +142,7 @@ static void __exit exit_gfs2_fs(void) unregister_filesystem(&gfs2_fs_type); unregister_filesystem(&gfs2meta_fs_type); + kmem_cache_destroy(gfs2_rgrpd_cachep); kmem_cache_destroy(gfs2_bufdata_cachep); kmem_cache_destroy(gfs2_inode_cachep); kmem_cache_destroy(gfs2_glock_cachep); diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c index ac772b6..7523999 100644 --- a/fs/gfs2/ops_address.c +++ b/fs/gfs2/ops_address.c @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -104,11 +104,9 @@ static int gfs2_writepage_common(struct page *page, loff_t i_size = i_size_read(inode); pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT; unsigned offset; - int ret = -EIO; if (gfs2_assert_withdraw(sdp, gfs2_glock_is_held_excl(ip->i_gl))) goto out; - ret = 0; if (current->journal_info) goto redirty; /* Is the page fully outside i_size? (truncate in progress) */ diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index f4842f2..f97a8b8 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -30,7 +30,6 @@ #include "glock.h" #include "glops.h" #include "inode.h" -#include "lm.h" #include "log.h" #include "meta_io.h" #include "quota.h" @@ -596,6 +595,36 @@ static int gfs2_setlease(struct file *file, long arg, struct file_lock **fl) return generic_setlease(file, arg, fl); } +static int gfs2_lm_plock_get(struct gfs2_sbd *sdp, struct lm_lockname *name, + struct file *file, struct file_lock *fl) +{ + int error = -EIO; + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + error = sdp->sd_lockstruct.ls_ops->lm_plock_get( + sdp->sd_lockstruct.ls_lockspace, name, file, fl); + return error; +} + +static int gfs2_lm_plock(struct gfs2_sbd *sdp, struct lm_lockname *name, + struct file *file, int cmd, struct file_lock *fl) +{ + int error = -EIO; + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + error = sdp->sd_lockstruct.ls_ops->lm_plock( + sdp->sd_lockstruct.ls_lockspace, name, file, cmd, fl); + return error; +} + +static int gfs2_lm_punlock(struct gfs2_sbd *sdp, struct lm_lockname *name, + struct file *file, struct file_lock *fl) +{ + int error = -EIO; + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + error = sdp->sd_lockstruct.ls_ops->lm_punlock( + sdp->sd_lockstruct.ls_lockspace, name, file, fl); + return error; +} + /** * gfs2_lock - acquire/release a posix lock on a file * @file: the file pointer diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 4bee6aa..c4b7a21 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -26,7 +26,6 @@ #include "glock.h" #include "glops.h" #include "inode.h" -#include "lm.h" #include "mount.h" #include "ops_fstype.h" #include "ops_dentry.h" @@ -363,6 +362,13 @@ static int map_journal_extents(struct gfs2_sbd *sdp) return rc; } +static void gfs2_lm_others_may_mount(struct gfs2_sbd *sdp) +{ + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + sdp->sd_lockstruct.ls_ops->lm_others_may_mount( + sdp->sd_lockstruct.ls_lockspace); +} + static int init_journal(struct gfs2_sbd *sdp, int undo) { struct gfs2_holder ji_gh; @@ -542,7 +548,7 @@ static int init_inodes(struct gfs2_sbd *sdp, int undo) } ip = GFS2_I(sdp->sd_rindex); set_bit(GLF_STICKY, &ip->i_gl->gl_flags); - sdp->sd_rindex_vn = ip->i_gl->gl_vn - 1; + sdp->sd_rindex_uptodate = 0; /* Read in the quota inode */ sdp->sd_quota_inode = gfs2_lookup_simple(sdp->sd_master_dir, "quota"); @@ -705,6 +711,69 @@ fail: } /** + * gfs2_lm_mount - mount a locking protocol + * @sdp: the filesystem + * @args: mount arguements + * @silent: if 1, don't complain if the FS isn't a GFS2 fs + * + * Returns: errno + */ + +static int gfs2_lm_mount(struct gfs2_sbd *sdp, int silent) +{ + char *proto = sdp->sd_proto_name; + char *table = sdp->sd_table_name; + int flags = 0; + int error; + + if (sdp->sd_args.ar_spectator) + flags |= LM_MFLAG_SPECTATOR; + + fs_info(sdp, "Trying to join cluster \"%s\", \"%s\"\n", proto, table); + + error = gfs2_mount_lockproto(proto, table, sdp->sd_args.ar_hostdata, + gfs2_glock_cb, sdp, + GFS2_MIN_LVB_SIZE, flags, + &sdp->sd_lockstruct, &sdp->sd_kobj); + if (error) { + fs_info(sdp, "can't mount proto=%s, table=%s, hostdata=%s\n", + proto, table, sdp->sd_args.ar_hostdata); + goto out; + } + + if (gfs2_assert_warn(sdp, sdp->sd_lockstruct.ls_lockspace) || + gfs2_assert_warn(sdp, sdp->sd_lockstruct.ls_ops) || + gfs2_assert_warn(sdp, sdp->sd_lockstruct.ls_lvb_size >= + GFS2_MIN_LVB_SIZE)) { + gfs2_unmount_lockproto(&sdp->sd_lockstruct); + goto out; + } + + if (sdp->sd_args.ar_spectator) + snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s.s", table); + else + snprintf(sdp->sd_fsname, GFS2_FSNAME_LEN, "%s.%u", table, + sdp->sd_lockstruct.ls_jid); + + fs_info(sdp, "Joined cluster. Now mounting FS...\n"); + + if ((sdp->sd_lockstruct.ls_flags & LM_LSFLAG_LOCAL) && + !sdp->sd_args.ar_ignore_local_fs) { + sdp->sd_args.ar_localflocks = 1; + sdp->sd_args.ar_localcaching = 1; + } + +out: + return error; +} + +void gfs2_lm_unmount(struct gfs2_sbd *sdp) +{ + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + gfs2_unmount_lockproto(&sdp->sd_lockstruct); +} + +/** * fill_super - Read in superblock * @sb: The VFS superblock * @data: Mount options diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c index e874129..301c945 100644 --- a/fs/gfs2/ops_inode.c +++ b/fs/gfs2/ops_inode.c @@ -981,8 +981,9 @@ static int setattr_chown(struct inode *inode, struct iattr *attr) brelse(dibh); if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { - gfs2_quota_change(ip, -ip->i_di.di_blocks, ouid, ogid); - gfs2_quota_change(ip, ip->i_di.di_blocks, nuid, ngid); + u64 blocks = gfs2_get_inode_blocks(&ip->i_inode); + gfs2_quota_change(ip, -blocks, ouid, ogid); + gfs2_quota_change(ip, blocks, nuid, ngid); } out_end_trans: diff --git a/fs/gfs2/ops_super.c b/fs/gfs2/ops_super.c index 5e52421..2278c68 100644 --- a/fs/gfs2/ops_super.c +++ b/fs/gfs2/ops_super.c @@ -25,7 +25,6 @@ #include "incore.h" #include "glock.h" #include "inode.h" -#include "lm.h" #include "log.h" #include "mount.h" #include "ops_super.h" diff --git a/fs/gfs2/recovery.c b/fs/gfs2/recovery.c index 6fb07d6..b17d3b8 100644 --- a/fs/gfs2/recovery.c +++ b/fs/gfs2/recovery.c @@ -20,7 +20,6 @@ #include "bmap.h" #include "glock.h" #include "glops.h" -#include "lm.h" #include "lops.h" #include "meta_io.h" #include "recovery.h" @@ -425,6 +424,16 @@ static int clean_journal(struct gfs2_jdesc *jd, struct gfs2_log_header_host *hea return error; } + +static void gfs2_lm_recovery_done(struct gfs2_sbd *sdp, unsigned int jid, + unsigned int message) +{ + if (likely(!test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) + sdp->sd_lockstruct.ls_ops->lm_recovery_done( + sdp->sd_lockstruct.ls_lockspace, jid, message); +} + + /** * gfs2_recover_journal - recovery a given journal * @jd: the struct gfs2_jdesc describing the journal diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 3552110..77eba0a 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2007 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -53,7 +53,8 @@ static const char valid_change[16] = { }; static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, - unsigned char old_state, unsigned char new_state); + unsigned char old_state, unsigned char new_state, + unsigned int *n); /** * gfs2_setbit - Set a bit in the bitmaps @@ -64,26 +65,32 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, * */ -static void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, - unsigned int buflen, u32 block, - unsigned char new_state) +static inline void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buf1, + unsigned char *buf2, unsigned int offset, + unsigned int buflen, u32 block, + unsigned char new_state) { - unsigned char *byte, *end, cur_state; - unsigned int bit; + unsigned char *byte1, *byte2, *end, cur_state; + const unsigned int bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE; - byte = buffer + (block / GFS2_NBBY); - bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE; - end = buffer + buflen; + byte1 = buf1 + offset + (block / GFS2_NBBY); + end = buf1 + offset + buflen; - gfs2_assert(rgd->rd_sbd, byte < end); + BUG_ON(byte1 >= end); - cur_state = (*byte >> bit) & GFS2_BIT_MASK; + cur_state = (*byte1 >> bit) & GFS2_BIT_MASK; - if (valid_change[new_state * 4 + cur_state]) { - *byte ^= cur_state << bit; - *byte |= new_state << bit; - } else + if (unlikely(!valid_change[new_state * 4 + cur_state])) { gfs2_consist_rgrpd(rgd); + return; + } + *byte1 ^= (cur_state ^ new_state) << bit; + + if (buf2) { + byte2 = buf2 + offset + (block / GFS2_NBBY); + cur_state = (*byte2 >> bit) & GFS2_BIT_MASK; + *byte2 ^= (cur_state ^ new_state) << bit; + } } /** @@ -94,10 +101,12 @@ static void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, * */ -static unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, - unsigned int buflen, u32 block) +static inline unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, + const unsigned char *buffer, + unsigned int buflen, u32 block) { - unsigned char *byte, *end, cur_state; + const unsigned char *byte, *end; + unsigned char cur_state; unsigned int bit; byte = buffer + (block / GFS2_NBBY); @@ -126,23 +135,24 @@ static unsigned char gfs2_testbit(struct gfs2_rgrpd *rgd, unsigned char *buffer, * Return: the block number (bitmap buffer scope) that was found */ -static u32 gfs2_bitfit(unsigned char *buffer, unsigned int buflen, u32 goal, - unsigned char old_state) +static u32 gfs2_bitfit(const u8 *buffer, unsigned int buflen, u32 goal, + u8 old_state) { - unsigned char *byte; + const u8 *byte; u32 blk = goal; unsigned int bit, bitlong; - unsigned long *plong, plong55; + const unsigned long *plong; +#if BITS_PER_LONG == 32 + const unsigned long plong55 = 0x55555555; +#else + const unsigned long plong55 = 0x5555555555555555; +#endif byte = buffer + (goal / GFS2_NBBY); - plong = (unsigned long *)(buffer + (goal / GFS2_NBBY)); + plong = (const unsigned long *)(buffer + (goal / GFS2_NBBY)); bit = (goal % GFS2_NBBY) * GFS2_BIT_SIZE; bitlong = bit; -#if BITS_PER_LONG == 32 - plong55 = 0x55555555; -#else - plong55 = 0x5555555555555555; -#endif + while (byte < buffer + buflen) { if (bitlong == 0 && old_state == 0 && *plong == plong55) { @@ -179,14 +189,14 @@ static u32 gfs2_bitfit(unsigned char *buffer, unsigned int buflen, u32 goal, * Returns: The number of bits */ -static u32 gfs2_bitcount(struct gfs2_rgrpd *rgd, unsigned char *buffer, - unsigned int buflen, unsigned char state) +static u32 gfs2_bitcount(struct gfs2_rgrpd *rgd, const u8 *buffer, + unsigned int buflen, u8 state) { - unsigned char *byte = buffer; - unsigned char *end = buffer + buflen; - unsigned char state1 = state << 2; - unsigned char state2 = state << 4; - unsigned char state3 = state << 6; + const u8 *byte = buffer; + const u8 *end = buffer + buflen; + const u8 state1 = state << 2; + const u8 state2 = state << 4; + const u8 state3 = state << 6; u32 count = 0; for (; byte < end; byte++) { @@ -353,7 +363,7 @@ static void clear_rgrpdi(struct gfs2_sbd *sdp) } kfree(rgd->rd_bits); - kfree(rgd); + kmem_cache_free(gfs2_rgrpd_cachep, rgd); } } @@ -516,7 +526,7 @@ static int read_rindex_entry(struct gfs2_inode *ip, return error; } - rgd = kzalloc(sizeof(struct gfs2_rgrpd), GFP_NOFS); + rgd = kmem_cache_zalloc(gfs2_rgrpd_cachep, GFP_NOFS); error = -ENOMEM; if (!rgd) return error; @@ -539,7 +549,7 @@ static int read_rindex_entry(struct gfs2_inode *ip, return error; rgd->rd_gl->gl_object = rgd; - rgd->rd_rg_vn = rgd->rd_gl->gl_vn - 1; + rgd->rd_flags &= ~GFS2_RDF_UPTODATE; rgd->rd_flags |= GFS2_RDF_CHECK; return error; } @@ -575,7 +585,7 @@ static int gfs2_ri_update(struct gfs2_inode *ip) } } - sdp->sd_rindex_vn = ip->i_gl->gl_vn; + sdp->sd_rindex_uptodate = 1; return 0; } @@ -609,7 +619,7 @@ static int gfs2_ri_update_special(struct gfs2_inode *ip) } } - sdp->sd_rindex_vn = ip->i_gl->gl_vn; + sdp->sd_rindex_uptodate = 1; return 0; } @@ -642,9 +652,9 @@ int gfs2_rindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ri_gh) return error; /* Read new copy from disk if we don't have the latest */ - if (sdp->sd_rindex_vn != gl->gl_vn) { + if (!sdp->sd_rindex_uptodate) { mutex_lock(&sdp->sd_rindex_mutex); - if (sdp->sd_rindex_vn != gl->gl_vn) { + if (!sdp->sd_rindex_uptodate) { error = gfs2_ri_update(ip); if (error) gfs2_glock_dq_uninit(ri_gh); @@ -655,21 +665,31 @@ int gfs2_rindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ri_gh) return error; } -static void gfs2_rgrp_in(struct gfs2_rgrp_host *rg, const void *buf) +static void gfs2_rgrp_in(struct gfs2_rgrpd *rgd, const void *buf) { const struct gfs2_rgrp *str = buf; + struct gfs2_rgrp_host *rg = &rgd->rd_rg; + u32 rg_flags; - rg->rg_flags = be32_to_cpu(str->rg_flags); + rg_flags = be32_to_cpu(str->rg_flags); + if (rg_flags & GFS2_RGF_NOALLOC) + rgd->rd_flags |= GFS2_RDF_NOALLOC; + else + rgd->rd_flags &= ~GFS2_RDF_NOALLOC; rg->rg_free = be32_to_cpu(str->rg_free); rg->rg_dinodes = be32_to_cpu(str->rg_dinodes); rg->rg_igeneration = be64_to_cpu(str->rg_igeneration); } -static void gfs2_rgrp_out(const struct gfs2_rgrp_host *rg, void *buf) +static void gfs2_rgrp_out(struct gfs2_rgrpd *rgd, void *buf) { struct gfs2_rgrp *str = buf; + struct gfs2_rgrp_host *rg = &rgd->rd_rg; + u32 rg_flags = 0; - str->rg_flags = cpu_to_be32(rg->rg_flags); + if (rgd->rd_flags & GFS2_RDF_NOALLOC) + rg_flags |= GFS2_RGF_NOALLOC; + str->rg_flags = cpu_to_be32(rg_flags); str->rg_free = cpu_to_be32(rg->rg_free); str->rg_dinodes = cpu_to_be32(rg->rg_dinodes); str->__pad = cpu_to_be32(0); @@ -726,9 +746,9 @@ int gfs2_rgrp_bh_get(struct gfs2_rgrpd *rgd) } } - if (rgd->rd_rg_vn != gl->gl_vn) { - gfs2_rgrp_in(&rgd->rd_rg, (rgd->rd_bits[0].bi_bh)->b_data); - rgd->rd_rg_vn = gl->gl_vn; + if (!(rgd->rd_flags & GFS2_RDF_UPTODATE)) { + gfs2_rgrp_in(rgd, (rgd->rd_bits[0].bi_bh)->b_data); + rgd->rd_flags |= GFS2_RDF_UPTODATE; } spin_lock(&sdp->sd_rindex_spin); @@ -840,7 +860,7 @@ static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al) struct gfs2_sbd *sdp = rgd->rd_sbd; int ret = 0; - if (rgd->rd_rg.rg_flags & GFS2_RGF_NOALLOC) + if (rgd->rd_flags & GFS2_RDF_NOALLOC) return 0; spin_lock(&sdp->sd_rindex_spin); @@ -866,13 +886,15 @@ static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked) u32 goal = 0, block; u64 no_addr; struct gfs2_sbd *sdp = rgd->rd_sbd; + unsigned int n; for(;;) { if (goal >= rgd->rd_data) break; down_write(&sdp->sd_log_flush_lock); + n = 1; block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED, - GFS2_BLKST_UNLINKED); + GFS2_BLKST_UNLINKED, &n); up_write(&sdp->sd_log_flush_lock); if (block == BFITNOENT) break; @@ -904,24 +926,20 @@ static struct inode *try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked) static struct gfs2_rgrpd *recent_rgrp_first(struct gfs2_sbd *sdp, u64 rglast) { - struct gfs2_rgrpd *rgd = NULL; + struct gfs2_rgrpd *rgd; spin_lock(&sdp->sd_rindex_spin); - if (list_empty(&sdp->sd_rindex_recent_list)) - goto out; - - if (!rglast) - goto first; - - list_for_each_entry(rgd, &sdp->sd_rindex_recent_list, rd_recent) { - if (rgd->rd_addr == rglast) - goto out; + if (rglast) { + list_for_each_entry(rgd, &sdp->sd_rindex_recent_list, rd_recent) { + if (rgrp_contains_block(rgd, rglast)) + goto out; + } } - -first: - rgd = list_entry(sdp->sd_rindex_recent_list.next, struct gfs2_rgrpd, - rd_recent); + rgd = NULL; + if (!list_empty(&sdp->sd_rindex_recent_list)) + rgd = list_entry(sdp->sd_rindex_recent_list.next, + struct gfs2_rgrpd, rd_recent); out: spin_unlock(&sdp->sd_rindex_spin); return rgd; @@ -1067,7 +1085,7 @@ static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked) /* Try recently successful rgrps */ - rgd = recent_rgrp_first(sdp, ip->i_last_rg_alloc); + rgd = recent_rgrp_first(sdp, ip->i_goal); while (rgd) { rg_locked = 0; @@ -1151,8 +1169,6 @@ static struct inode *get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked) } out: - ip->i_last_rg_alloc = rgd->rd_addr; - if (begin) { recent_rgrp_add(rgd); rgd = gfs2_rgrpd_get_next(rgd); @@ -1275,6 +1291,7 @@ unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block) * @goal: the goal block within the RG (start here to search for avail block) * @old_state: GFS2_BLKST_XXX the before-allocation state to find * @new_state: GFS2_BLKST_XXX the after-allocation block state + * @n: The extent length * * Walk rgrp's bitmap to find bits that represent a block in @old_state. * Add the found bitmap buffer to the transaction. @@ -1290,13 +1307,17 @@ unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block) */ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, - unsigned char old_state, unsigned char new_state) + unsigned char old_state, unsigned char new_state, + unsigned int *n) { struct gfs2_bitmap *bi = NULL; - u32 length = rgd->rd_length; + const u32 length = rgd->rd_length; u32 blk = 0; unsigned int buf, x; + const unsigned int elen = *n; + const u8 *buffer; + *n = 0; /* Find bitmap block that contains bits for goal block */ for (buf = 0; buf < length; buf++) { bi = rgd->rd_bits + buf; @@ -1317,12 +1338,11 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, for (x = 0; x <= length; x++) { /* The GFS2_BLKST_UNLINKED state doesn't apply to the clone bitmaps, so we must search the originals for that. */ + buffer = bi->bi_bh->b_data + bi->bi_offset; if (old_state != GFS2_BLKST_UNLINKED && bi->bi_clone) - blk = gfs2_bitfit(bi->bi_clone + bi->bi_offset, - bi->bi_len, goal, old_state); - else - blk = gfs2_bitfit(bi->bi_bh->b_data + bi->bi_offset, - bi->bi_len, goal, old_state); + buffer = bi->bi_clone + bi->bi_offset; + + blk = gfs2_bitfit(buffer, bi->bi_len, goal, old_state); if (blk != BFITNOENT) break; @@ -1333,12 +1353,21 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, } if (blk != BFITNOENT && old_state != new_state) { + *n = 1; gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); - gfs2_setbit(rgd, bi->bi_bh->b_data + bi->bi_offset, + gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset, bi->bi_len, blk, new_state); - if (bi->bi_clone) - gfs2_setbit(rgd, bi->bi_clone + bi->bi_offset, - bi->bi_len, blk, new_state); + while(*n < elen) { + goal++; + if (goal >= (bi->bi_len / GFS2_NBBY)) + break; + if (gfs2_testbit(rgd, buffer, bi->bi_len, goal) != + GFS2_BLKST_FREE) + break; + (*n)++; + gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, + bi->bi_offset, bi->bi_len, blk, new_state); + } } return (blk == BFITNOENT) ? blk : (bi->bi_start * GFS2_NBBY) + blk; @@ -1393,7 +1422,7 @@ static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, bi->bi_len); } gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); - gfs2_setbit(rgd, bi->bi_bh->b_data + bi->bi_offset, + gfs2_setbit(rgd, bi->bi_bh->b_data, NULL, bi->bi_offset, bi->bi_len, buf_blk, new_state); } @@ -1401,58 +1430,13 @@ static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, } /** - * gfs2_alloc_data - Allocate a data block - * @ip: the inode to allocate the data block for - * - * Returns: the allocated block - */ - -u64 gfs2_alloc_data(struct gfs2_inode *ip) -{ - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al = ip->i_alloc; - struct gfs2_rgrpd *rgd = al->al_rgd; - u32 goal, blk; - u64 block; - - if (rgrp_contains_block(rgd, ip->i_di.di_goal_data)) - goal = ip->i_di.di_goal_data - rgd->rd_data0; - else - goal = rgd->rd_last_alloc_data; - - blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED); - BUG_ON(blk == BFITNOENT); - rgd->rd_last_alloc_data = blk; - - block = rgd->rd_data0 + blk; - ip->i_di.di_goal_data = block; - - gfs2_assert_withdraw(sdp, rgd->rd_rg.rg_free); - rgd->rd_rg.rg_free--; - - gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(&rgd->rd_rg, rgd->rd_bits[0].bi_bh->b_data); - - al->al_alloced++; - - gfs2_statfs_change(sdp, 0, -1, 0); - gfs2_quota_change(ip, +1, ip->i_inode.i_uid, ip->i_inode.i_gid); - - spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone--; - spin_unlock(&sdp->sd_rindex_spin); - - return block; -} - -/** - * gfs2_alloc_meta - Allocate a metadata block - * @ip: the inode to allocate the metadata block for + * gfs2_alloc_block - Allocate a block + * @ip: the inode to allocate the block for * * Returns: the allocated block */ -u64 gfs2_alloc_meta(struct gfs2_inode *ip) +u64 gfs2_alloc_block(struct gfs2_inode *ip, unsigned int *n) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_alloc *al = ip->i_alloc; @@ -1460,32 +1444,31 @@ u64 gfs2_alloc_meta(struct gfs2_inode *ip) u32 goal, blk; u64 block; - if (rgrp_contains_block(rgd, ip->i_di.di_goal_meta)) - goal = ip->i_di.di_goal_meta - rgd->rd_data0; + if (rgrp_contains_block(rgd, ip->i_goal)) + goal = ip->i_goal - rgd->rd_data0; else - goal = rgd->rd_last_alloc_meta; + goal = rgd->rd_last_alloc; - blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED); + blk = rgblk_search(rgd, goal, GFS2_BLKST_FREE, GFS2_BLKST_USED, n); BUG_ON(blk == BFITNOENT); - rgd->rd_last_alloc_meta = blk; + rgd->rd_last_alloc = blk; block = rgd->rd_data0 + blk; - ip->i_di.di_goal_meta = block; + ip->i_goal = block; - gfs2_assert_withdraw(sdp, rgd->rd_rg.rg_free); - rgd->rd_rg.rg_free--; + gfs2_assert_withdraw(sdp, rgd->rd_rg.rg_free >= *n); + rgd->rd_rg.rg_free -= *n; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(&rgd->rd_rg, rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); - al->al_alloced++; + al->al_alloced += *n; - gfs2_statfs_change(sdp, 0, -1, 0); - gfs2_quota_change(ip, +1, ip->i_inode.i_uid, ip->i_inode.i_gid); - gfs2_trans_add_unrevoke(sdp, block); + gfs2_statfs_change(sdp, 0, -*n, 0); + gfs2_quota_change(ip, *n, ip->i_inode.i_uid, ip->i_inode.i_gid); spin_lock(&sdp->sd_rindex_spin); - rgd->rd_free_clone--; + rgd->rd_free_clone -= *n; spin_unlock(&sdp->sd_rindex_spin); return block; @@ -1505,12 +1488,13 @@ u64 gfs2_alloc_di(struct gfs2_inode *dip, u64 *generation) struct gfs2_rgrpd *rgd = al->al_rgd; u32 blk; u64 block; + unsigned int n = 1; - blk = rgblk_search(rgd, rgd->rd_last_alloc_meta, - GFS2_BLKST_FREE, GFS2_BLKST_DINODE); + blk = rgblk_search(rgd, rgd->rd_last_alloc, + GFS2_BLKST_FREE, GFS2_BLKST_DINODE, &n); BUG_ON(blk == BFITNOENT); - rgd->rd_last_alloc_meta = blk; + rgd->rd_last_alloc = blk; block = rgd->rd_data0 + blk; @@ -1519,12 +1503,12 @@ u64 gfs2_alloc_di(struct gfs2_inode *dip, u64 *generation) rgd->rd_rg.rg_dinodes++; *generation = rgd->rd_rg.rg_igeneration++; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(&rgd->rd_rg, rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); al->al_alloced++; gfs2_statfs_change(sdp, 0, -1, +1); - gfs2_trans_add_unrevoke(sdp, block); + gfs2_trans_add_unrevoke(sdp, block, 1); spin_lock(&sdp->sd_rindex_spin); rgd->rd_free_clone--; @@ -1553,7 +1537,7 @@ void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) rgd->rd_rg.rg_free += blen; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(&rgd->rd_rg, rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_trans_add_rg(rgd); @@ -1581,7 +1565,7 @@ void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen) rgd->rd_rg.rg_free += blen; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(&rgd->rd_rg, rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_trans_add_rg(rgd); @@ -1601,7 +1585,7 @@ void gfs2_unlink_di(struct inode *inode) if (!rgd) return; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(&rgd->rd_rg, rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_trans_add_rg(rgd); } @@ -1621,7 +1605,7 @@ static void gfs2_free_uninit_di(struct gfs2_rgrpd *rgd, u64 blkno) rgd->rd_rg.rg_free++; gfs2_trans_add_bh(rgd->rd_gl, rgd->rd_bits[0].bi_bh, 1); - gfs2_rgrp_out(&rgd->rd_rg, rgd->rd_bits[0].bi_bh->b_data); + gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_statfs_change(sdp, 0, +1, -1); gfs2_trans_add_rg(rgd); @@ -1699,8 +1683,7 @@ void gfs2_rlist_add(struct gfs2_sbd *sdp, struct gfs2_rgrp_list *rlist, * */ -void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state, - int flags) +void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state) { unsigned int x; @@ -1708,7 +1691,7 @@ void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state, GFP_NOFS | __GFP_NOFAIL); for (x = 0; x < rlist->rl_rgrps; x++) gfs2_holder_init(rlist->rl_rgd[x]->rd_gl, - state, flags, + state, 0, &rlist->rl_ghs[x]); } diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h index 149bb16..3181c7e 100644 --- a/fs/gfs2/rgrp.h +++ b/fs/gfs2/rgrp.h @@ -1,6 +1,6 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions @@ -46,8 +46,7 @@ void gfs2_inplace_release(struct gfs2_inode *ip); unsigned char gfs2_get_block_type(struct gfs2_rgrpd *rgd, u64 block); -u64 gfs2_alloc_data(struct gfs2_inode *ip); -u64 gfs2_alloc_meta(struct gfs2_inode *ip); +u64 gfs2_alloc_block(struct gfs2_inode *ip, unsigned int *n); u64 gfs2_alloc_di(struct gfs2_inode *ip, u64 *generation); void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen); @@ -64,8 +63,7 @@ struct gfs2_rgrp_list { void gfs2_rlist_add(struct gfs2_sbd *sdp, struct gfs2_rgrp_list *rlist, u64 block); -void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state, - int flags); +void gfs2_rlist_alloc(struct gfs2_rgrp_list *rlist, unsigned int state); void gfs2_rlist_free(struct gfs2_rgrp_list *rlist); u64 gfs2_ri_total(struct gfs2_sbd *sdp); diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index ef0562c..88497b0 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -316,6 +316,7 @@ int gfs2_read_sb(struct gfs2_sbd *sdp, struct gfs2_glock *gl, int silent) sdp->sd_heightsize[x] = space; } sdp->sd_max_height = x; + sdp->sd_heightsize[x] = ~0; gfs2_assert(sdp, sdp->sd_max_height <= GFS2_MAX_META_HEIGHT); sdp->sd_jheightsize[0] = sdp->sd_sb.sb_bsize - @@ -334,6 +335,7 @@ int gfs2_read_sb(struct gfs2_sbd *sdp, struct gfs2_glock *gl, int silent) sdp->sd_jheightsize[x] = space; } sdp->sd_max_jheight = x; + sdp->sd_jheightsize[x] = ~0; gfs2_assert(sdp, sdp->sd_max_jheight <= GFS2_MAX_META_HEIGHT); return 0; diff --git a/fs/gfs2/super.h b/fs/gfs2/super.h index 60a870e..44361ec 100644 --- a/fs/gfs2/super.h +++ b/fs/gfs2/super.h @@ -17,6 +17,7 @@ void gfs2_tune_init(struct gfs2_tune *gt); int gfs2_check_sb(struct gfs2_sbd *sdp, struct gfs2_sb_host *sb, int silent); int gfs2_read_sb(struct gfs2_sbd *sdp, struct gfs2_glock *gl, int silent); int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector); +void gfs2_lm_unmount(struct gfs2_sbd *sdp); static inline unsigned int gfs2_jindex_size(struct gfs2_sbd *sdp) { diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index eaa3b7b..9ab9fc8 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -20,7 +20,6 @@ #include "gfs2.h" #include "incore.h" -#include "lm.h" #include "sys.h" #include "super.h" #include "glock.h" @@ -328,15 +327,9 @@ static ssize_t name##_show(struct gfs2_sbd *sdp, char *buf) \ } \ static struct counters_attr counters_attr_##name = __ATTR_RO(name) -COUNTERS_ATTR(glock_count, "%u\n"); -COUNTERS_ATTR(glock_held_count, "%u\n"); -COUNTERS_ATTR(inode_count, "%u\n"); COUNTERS_ATTR(reclaimed, "%u\n"); static struct attribute *counters_attrs[] = { - &counters_attr_glock_count.attr, - &counters_attr_glock_held_count.attr, - &counters_attr_inode_count.attr, &counters_attr_reclaimed.attr, NULL, }; diff --git a/fs/gfs2/trans.c b/fs/gfs2/trans.c index 73e5d92..f677b8a 100644 --- a/fs/gfs2/trans.c +++ b/fs/gfs2/trans.c @@ -146,30 +146,25 @@ void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd) lops_add(sdp, &bd->bd_le); } -void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno) +void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len) { - struct gfs2_bufdata *bd; - int found = 0; + struct gfs2_bufdata *bd, *tmp; + struct gfs2_trans *tr = current->journal_info; + unsigned int n = len; gfs2_log_lock(sdp); - - list_for_each_entry(bd, &sdp->sd_log_le_revoke, bd_le.le_list) { - if (bd->bd_blkno == blkno) { + list_for_each_entry_safe(bd, tmp, &sdp->sd_log_le_revoke, bd_le.le_list) { + if ((bd->bd_blkno >= blkno) && (bd->bd_blkno < (blkno + len))) { list_del_init(&bd->bd_le.le_list); gfs2_assert_withdraw(sdp, sdp->sd_log_num_revoke); sdp->sd_log_num_revoke--; - found = 1; - break; + kmem_cache_free(gfs2_bufdata_cachep, bd); + tr->tr_num_revoke_rm++; + if (--n == 0) + break; } } - gfs2_log_unlock(sdp); - - if (found) { - struct gfs2_trans *tr = current->journal_info; - kmem_cache_free(gfs2_bufdata_cachep, bd); - tr->tr_num_revoke_rm++; - } } void gfs2_trans_add_rg(struct gfs2_rgrpd *rgd) diff --git a/fs/gfs2/trans.h b/fs/gfs2/trans.h index e826f0d..edf9d4b 100644 --- a/fs/gfs2/trans.h +++ b/fs/gfs2/trans.h @@ -32,7 +32,7 @@ void gfs2_trans_end(struct gfs2_sbd *sdp); void gfs2_trans_add_bh(struct gfs2_glock *gl, struct buffer_head *bh, int meta); void gfs2_trans_add_revoke(struct gfs2_sbd *sdp, struct gfs2_bufdata *bd); -void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno); +void gfs2_trans_add_unrevoke(struct gfs2_sbd *sdp, u64 blkno, unsigned int len); void gfs2_trans_add_rg(struct gfs2_rgrpd *rgd); #endif /* __TRANS_DOT_H__ */ diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c index 424a077..d31e355 100644 --- a/fs/gfs2/util.c +++ b/fs/gfs2/util.c @@ -19,12 +19,12 @@ #include "gfs2.h" #include "incore.h" #include "glock.h" -#include "lm.h" #include "util.h" struct kmem_cache *gfs2_glock_cachep __read_mostly; struct kmem_cache *gfs2_inode_cachep __read_mostly; struct kmem_cache *gfs2_bufdata_cachep __read_mostly; +struct kmem_cache *gfs2_rgrpd_cachep __read_mostly; void gfs2_assert_i(struct gfs2_sbd *sdp) { @@ -32,6 +32,28 @@ void gfs2_assert_i(struct gfs2_sbd *sdp) sdp->sd_fsname); } +int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...) +{ + va_list args; + + if (test_and_set_bit(SDF_SHUTDOWN, &sdp->sd_flags)) + return 0; + + va_start(args, fmt); + vprintk(fmt, args); + va_end(args); + + fs_err(sdp, "about to withdraw this file system\n"); + BUG_ON(sdp->sd_args.ar_debug); + + fs_err(sdp, "telling LM to withdraw\n"); + gfs2_withdraw_lockproto(&sdp->sd_lockstruct); + fs_err(sdp, "withdrawn\n"); + dump_stack(); + + return -1; +} + /** * gfs2_assert_withdraw_i - Cause the machine to withdraw if @assertion is false * Returns: -1 if this call withdrew the machine, diff --git a/fs/gfs2/util.h b/fs/gfs2/util.h index 28938a4..509c5d6 100644 --- a/fs/gfs2/util.h +++ b/fs/gfs2/util.h @@ -147,6 +147,7 @@ gfs2_io_error_bh_i((sdp), (bh), __FUNCTION__, __FILE__, __LINE__); extern struct kmem_cache *gfs2_glock_cachep; extern struct kmem_cache *gfs2_inode_cachep; extern struct kmem_cache *gfs2_bufdata_cachep; +extern struct kmem_cache *gfs2_rgrpd_cachep; static inline unsigned int gfs2_tune_get_i(struct gfs2_tune *gt, unsigned int *p) @@ -163,6 +164,7 @@ gfs2_tune_get_i(&(sdp)->sd_tune, &(sdp)->sd_tune.field) void gfs2_icbit_munge(struct gfs2_sbd *sdp, unsigned char **bitmap, unsigned int bit, int new_value); +int gfs2_lm_withdraw(struct gfs2_sbd *sdp, char *fmt, ...); #endif /* __UTIL_DOT_H__ */