* David S. Miller (davem@caip.rutgers.edu), 1995
*/
-#include <linux/config.h>
#include <linux/time.h>
#include <linux/capability.h>
#include <linux/fs.h>
* The file system contains group descriptors which are located after the
* super block. Each descriptor contains the number of the bitmap block and
* the free blocks count in the block. The descriptors are loaded in memory
- * when a file system is mounted (see ext3_read_super).
+ * when a file system is mounted (see ext3_fill_super).
*/
#define in_range(b, first, len) ((b) >= (first) && (b) <= (first) + (len) - 1)
+/**
+ * ext3_get_group_desc() -- load group descriptor from disk
+ * @sb: super block
+ * @block_group: given block group
+ * @bh: pointer to the buffer head to store the block
+ * group descriptor
+ */
struct ext3_group_desc * ext3_get_group_desc(struct super_block * sb,
unsigned int block_group,
struct buffer_head ** bh)
return desc + offset;
}
-/*
- * Read the bitmap for a given block_group, reading into the specified
- * slot in the superblock's bitmap cache.
+static int ext3_valid_block_bitmap(struct super_block *sb,
+ struct ext3_group_desc *desc,
+ unsigned int block_group,
+ struct buffer_head *bh)
+{
+ ext3_grpblk_t offset;
+ ext3_grpblk_t next_zero_bit;
+ ext3_fsblk_t bitmap_blk;
+ ext3_fsblk_t group_first_block;
+
+ group_first_block = ext3_group_first_block_no(sb, block_group);
+
+ /* check whether block bitmap block number is set */
+ bitmap_blk = le32_to_cpu(desc->bg_block_bitmap);
+ offset = bitmap_blk - group_first_block;
+ if (!ext3_test_bit(offset, bh->b_data))
+ /* bad block bitmap */
+ goto err_out;
+
+ /* check whether the inode bitmap block number is set */
+ bitmap_blk = le32_to_cpu(desc->bg_inode_bitmap);
+ offset = bitmap_blk - group_first_block;
+ if (!ext3_test_bit(offset, bh->b_data))
+ /* bad block bitmap */
+ goto err_out;
+
+ /* check whether the inode table block number is set */
+ bitmap_blk = le32_to_cpu(desc->bg_inode_table);
+ offset = bitmap_blk - group_first_block;
+ next_zero_bit = ext3_find_next_zero_bit(bh->b_data,
+ offset + EXT3_SB(sb)->s_itb_per_group,
+ offset);
+ if (next_zero_bit >= offset + EXT3_SB(sb)->s_itb_per_group)
+ /* good bitmap for inode tables */
+ return 1;
+
+err_out:
+ ext3_error(sb, __func__,
+ "Invalid block bitmap - "
+ "block_group = %d, block = %lu",
+ block_group, bitmap_blk);
+ return 0;
+}
+
+/**
+ * read_block_bitmap()
+ * @sb: super block
+ * @block_group: given block group
+ *
+ * Read the bitmap for a given block_group,and validate the
+ * bits for block/inode/inode tables are set in the bitmaps
*
* Return buffer_head on success or NULL in case of failure.
*/
{
struct ext3_group_desc * desc;
struct buffer_head * bh = NULL;
+ ext3_fsblk_t bitmap_blk;
- desc = ext3_get_group_desc (sb, block_group, NULL);
+ desc = ext3_get_group_desc(sb, block_group, NULL);
if (!desc)
- goto error_out;
- bh = sb_bread(sb, le32_to_cpu(desc->bg_block_bitmap));
- if (!bh)
- ext3_error (sb, "read_block_bitmap",
+ return NULL;
+ bitmap_blk = le32_to_cpu(desc->bg_block_bitmap);
+ bh = sb_getblk(sb, bitmap_blk);
+ if (unlikely(!bh)) {
+ ext3_error(sb, __func__,
+ "Cannot read block bitmap - "
+ "block_group = %d, block_bitmap = %u",
+ block_group, le32_to_cpu(desc->bg_block_bitmap));
+ return NULL;
+ }
+ if (likely(bh_uptodate_or_lock(bh)))
+ return bh;
+
+ if (bh_submit_read(bh) < 0) {
+ brelse(bh);
+ ext3_error(sb, __func__,
"Cannot read block bitmap - "
"block_group = %d, block_bitmap = %u",
block_group, le32_to_cpu(desc->bg_block_bitmap));
-error_out:
+ return NULL;
+ }
+ ext3_valid_block_bitmap(sb, desc, block_group, bh);
+ /*
+ * file system mounted not to panic on error, continue with corrupt
+ * bitmap
+ */
return bh;
}
/*
* Operations include:
* dump, find, add, remove, is_empty, find_next_reservable_window, etc.
*
- * We use sorted double linked list for the per-filesystem reservation
- * window list. (like in vm_region).
+ * We use a red-black tree to represent per-filesystem reservation
+ * windows.
*
- * Initially, we keep those small operations in the abstract functions,
- * so later if we need a better searching tree than double linked-list,
- * we could easily switch to that without changing too much
- * code.
*/
-#if 0
+
+/**
+ * __rsv_window_dump() -- Dump the filesystem block allocation reservation map
+ * @rb_root: root of per-filesystem reservation rb tree
+ * @verbose: verbose mode
+ * @fn: function which wishes to dump the reservation map
+ *
+ * If verbose is turned on, it will print the whole block reservation
+ * windows(start, end). Otherwise, it will only print out the "bad" windows,
+ * those windows that overlap with their immediate neighbors.
+ */
+#if 1
static void __rsv_window_dump(struct rb_root *root, int verbose,
const char *fn)
{
printk("Block Allocation Reservation Windows Map (%s):\n", fn);
while (n) {
- rsv = list_entry(n, struct ext3_reserve_window_node, rsv_node);
+ rsv = rb_entry(n, struct ext3_reserve_window_node, rsv_node);
if (verbose)
printk("reservation window 0x%p "
- "start: %d, end: %d\n",
+ "start: %lu, end: %lu\n",
rsv, rsv->rsv_start, rsv->rsv_end);
if (rsv->rsv_start && rsv->rsv_start >= rsv->rsv_end) {
printk("Bad reservation %p (start >= end)\n",
prev = rsv;
}
printk("Window map complete.\n");
- if (bad)
- BUG();
+ BUG_ON(bad);
}
#define rsv_window_dump(root, verbose) \
- __rsv_window_dump((root), (verbose), __FUNCTION__)
+ __rsv_window_dump((root), (verbose), __func__)
#else
#define rsv_window_dump(root, verbose) do {} while (0)
#endif
+/**
+ * goal_in_my_reservation()
+ * @rsv: inode's reservation window
+ * @grp_goal: given goal block relative to the allocation block group
+ * @group: the current allocation block group
+ * @sb: filesystem super block
+ *
+ * Test if the given goal block (group relative) is within the file's
+ * own block reservation window range.
+ *
+ * If the reservation window is outside the goal allocation group, return 0;
+ * grp_goal (given goal block) could be -1, which means no specific
+ * goal block. In this case, always return 1.
+ * If the goal block is within the reservation window, return 1;
+ * otherwise, return 0;
+ */
static int
goal_in_my_reservation(struct ext3_reserve_window *rsv, ext3_grpblk_t grp_goal,
unsigned int group, struct super_block * sb)
ext3_fsblk_t group_first_block, group_last_block;
group_first_block = ext3_group_first_block_no(sb, group);
- group_last_block = group_first_block + EXT3_BLOCKS_PER_GROUP(sb) - 1;
+ group_last_block = group_first_block + (EXT3_BLOCKS_PER_GROUP(sb) - 1);
if ((rsv->_rsv_start > group_last_block) ||
(rsv->_rsv_end < group_first_block))
return 1;
}
-/*
+/**
+ * search_reserve_window()
+ * @rb_root: root of reservation tree
+ * @goal: target allocation block
+ *
* Find the reserved window which includes the goal, or the previous one
* if the goal is not in any window.
* Returns NULL if there are no windows or if all windows start after the goal.
return rsv;
}
+/**
+ * ext3_rsv_window_add() -- Insert a window to the block reservation rb tree.
+ * @sb: super block
+ * @rsv: reservation window to add
+ *
+ * Must be called with rsv_lock hold.
+ */
void ext3_rsv_window_add(struct super_block *sb,
struct ext3_reserve_window_node *rsv)
{
p = &(*p)->rb_left;
else if (start > this->rsv_end)
p = &(*p)->rb_right;
- else
+ else {
+ rsv_window_dump(root, 1);
BUG();
+ }
}
rb_link_node(node, parent, p);
rb_insert_color(node, root);
}
+/**
+ * ext3_rsv_window_remove() -- unlink a window from the reservation rb tree
+ * @sb: super block
+ * @rsv: reservation window to remove
+ *
+ * Mark the block reservation window as not allocated, and unlink it
+ * from the filesystem reservation window rb tree. Must be called with
+ * rsv_lock hold.
+ */
static void rsv_window_remove(struct super_block *sb,
struct ext3_reserve_window_node *rsv)
{
rb_erase(&rsv->rsv_node, &EXT3_SB(sb)->s_rsv_window_root);
}
+/*
+ * rsv_is_empty() -- Check if the reservation window is allocated.
+ * @rsv: given reservation window to check
+ *
+ * returns 1 if the end block is EXT3_RESERVE_WINDOW_NOT_ALLOCATED.
+ */
static inline int rsv_is_empty(struct ext3_reserve_window *rsv)
{
/* a valid reservation end block could not be 0 */
- return (rsv->_rsv_end == EXT3_RESERVE_WINDOW_NOT_ALLOCATED);
+ return rsv->_rsv_end == EXT3_RESERVE_WINDOW_NOT_ALLOCATED;
}
+
+/**
+ * ext3_init_block_alloc_info()
+ * @inode: file inode structure
+ *
+ * Allocate and initialize the reservation window structure, and
+ * link the window to the ext3 inode structure at last
+ *
+ * The reservation window structure is only dynamically allocated
+ * and linked to ext3 inode the first time the open file
+ * needs a new block. So, before every ext3_new_block(s) call, for
+ * regular files, we should check whether the reservation window
+ * structure exists or not. In the latter case, this function is called.
+ * Fail to do so will result in block reservation being turned off for that
+ * open file.
+ *
+ * This function is called from ext3_get_blocks_handle(), also called
+ * when setting the reservation window size through ioctl before the file
+ * is open for write (needs block allocation).
+ *
+ * Needs truncate_mutex protection prior to call this function.
+ */
void ext3_init_block_alloc_info(struct inode *inode)
{
struct ext3_inode_info *ei = EXT3_I(inode);
rsv->rsv_start = EXT3_RESERVE_WINDOW_NOT_ALLOCATED;
rsv->rsv_end = EXT3_RESERVE_WINDOW_NOT_ALLOCATED;
- /*
+ /*
* if filesystem is mounted with NORESERVATION, the goal
* reservation window size is set to zero to indicate
* block reservation is off
ei->i_block_alloc_info = block_i;
}
+/**
+ * ext3_discard_reservation()
+ * @inode: inode
+ *
+ * Discard(free) block reservation window on last file close, or truncate
+ * or at last iput().
+ *
+ * It is being called in three cases:
+ * ext3_release_file(): last writer close the file
+ * ext3_clear_inode(): last iput(), when nobody link to this file.
+ * ext3_truncate(): when the block indirect map is about to change.
+ *
+ */
void ext3_discard_reservation(struct inode *inode)
{
struct ext3_inode_info *ei = EXT3_I(inode);
}
}
-/* Free given blocks, update quota and i_blocks field */
+/**
+ * ext3_free_blocks_sb() -- Free given blocks and update quota
+ * @handle: handle to this transaction
+ * @sb: super block
+ * @block: start physcial block to free
+ * @count: number of blocks to free
+ * @pdquot_freed_blocks: pointer to quota
+ */
void ext3_free_blocks_sb(handle_t *handle, struct super_block *sb,
ext3_fsblk_t block, unsigned long count,
unsigned long *pdquot_freed_blocks)
in_range (block, le32_to_cpu(desc->bg_inode_table),
sbi->s_itb_per_group) ||
in_range (block + count - 1, le32_to_cpu(desc->bg_inode_table),
- sbi->s_itb_per_group))
+ sbi->s_itb_per_group)) {
ext3_error (sb, "ext3_free_blocks",
"Freeing blocks in system zones - "
"Block = "E3FSBLK", count = %lu",
block, count);
+ goto error_return;
+ }
/*
* We are about to start releasing blocks in the bitmap,
}
/* @@@ This prevents newly-allocated data from being
* freed and then reallocated within the same
- * transaction.
- *
+ * transaction.
+ *
* Ideally we would want to allow that to happen, but to
* do so requires making journal_forget() capable of
* revoking the queued write of a data block, which
* safe not to set the allocation bit in the committed
* bitmap, because we know that there is no outstanding
* activity on the buffer any more and so it is safe to
- * reallocate it.
+ * reallocate it.
*/
BUFFER_TRACE(bitmap_bh, "set in b_committed_data");
J_ASSERT_BH(bitmap_bh,
if (!ext3_clear_bit_atomic(sb_bgl_lock(sbi, block_group),
bit + i, bitmap_bh->b_data)) {
jbd_unlock_bh_state(bitmap_bh);
- ext3_error(sb, __FUNCTION__,
+ ext3_error(sb, __func__,
"bit already cleared for block "E3FSBLK,
block + i);
jbd_lock_bh_state(bitmap_bh);
jbd_unlock_bh_state(bitmap_bh);
spin_lock(sb_bgl_lock(sbi, block_group));
- desc->bg_free_blocks_count =
- cpu_to_le16(le16_to_cpu(desc->bg_free_blocks_count) +
- group_freed);
+ le16_add_cpu(&desc->bg_free_blocks_count, group_freed);
spin_unlock(sb_bgl_lock(sbi, block_group));
- percpu_counter_mod(&sbi->s_freeblocks_counter, count);
+ percpu_counter_add(&sbi->s_freeblocks_counter, count);
/* We dirtied the bitmap block */
BUFFER_TRACE(bitmap_bh, "dirtied bitmap block");
return;
}
-/* Free given blocks, update quota and i_blocks field */
+/**
+ * ext3_free_blocks() -- Free given blocks and update quota
+ * @handle: handle for this transaction
+ * @inode: inode
+ * @block: start physical block to free
+ * @count: number of blocks to count
+ */
void ext3_free_blocks(handle_t *handle, struct inode *inode,
ext3_fsblk_t block, unsigned long count)
{
return;
}
-/*
+/**
+ * ext3_test_allocatable()
+ * @nr: given allocation block group
+ * @bh: bufferhead contains the bitmap of the given block group
+ *
* For ext3 allocations, we must not reuse any blocks which are
* allocated in the bitmap buffer's "last committed data" copy. This
* prevents deletes from freeing up the page for reuse until we have
* data would allow the old block to be overwritten before the
* transaction committed (because we force data to disk before commit).
* This would lead to corruption if we crashed between overwriting the
- * data and committing the delete.
+ * data and committing the delete.
*
* @@@ We may want to make this allocation behaviour conditional on
* data-writes at some point, and disable it for metadata allocations or
return ret;
}
+/**
+ * bitmap_search_next_usable_block()
+ * @start: the starting block (group relative) of the search
+ * @bh: bufferhead contains the block group bitmap
+ * @maxblocks: the ending block (group relative) of the reservation
+ *
+ * The bitmap search --- search forward alternately through the actual
+ * bitmap on disk and the last-committed copy in journal, until we find a
+ * bit free in both bitmaps.
+ */
static ext3_grpblk_t
bitmap_search_next_usable_block(ext3_grpblk_t start, struct buffer_head *bh,
ext3_grpblk_t maxblocks)
ext3_grpblk_t next;
struct journal_head *jh = bh2jh(bh);
- /*
- * The bitmap search --- search forward alternately through the actual
- * bitmap and the last-committed copy until we find a bit free in
- * both
- */
while (start < maxblocks) {
next = ext3_find_next_zero_bit(bh->b_data, maxblocks, start);
if (next >= maxblocks)
jbd_lock_bh_state(bh);
if (jh->b_committed_data)
start = ext3_find_next_zero_bit(jh->b_committed_data,
- maxblocks, next);
+ maxblocks, next);
jbd_unlock_bh_state(bh);
}
return -1;
}
-/*
- * Find an allocatable block in a bitmap. We honour both the bitmap and
+/**
+ * find_next_usable_block()
+ * @start: the starting block (group relative) to find next
+ * allocatable block in bitmap.
+ * @bh: bufferhead contains the block group bitmap
+ * @maxblocks: the ending block (group relative) for the search
+ *
+ * Find an allocatable block in a bitmap. We honor both the bitmap and
* its last-committed copy (if that exists), and perform the "most
* appropriate allocation" algorithm of looking for a free block near
* the initial goal; then for a free byte somewhere in the bitmap; then
if (start > 0) {
/*
- * The goal was occupied; search forward for a free
+ * The goal was occupied; search forward for a free
* block within the next XX blocks.
*
* end_goal is more or less random, but it has to be
here = 0;
p = ((char *)bh->b_data) + (here >> 3);
- r = memscan(p, 0, (maxblocks - here + 7) >> 3);
+ r = memscan(p, 0, ((maxblocks + 7) >> 3) - (here >> 3));
next = (r - ((char *)bh->b_data)) << 3;
if (next < maxblocks && next >= start && ext3_test_allocatable(next, bh))
return here;
}
-/*
+/**
+ * claim_block()
+ * @block: the free block (group relative) to allocate
+ * @bh: the bufferhead containts the block group bitmap
+ *
* We think we can allocate this block in this bitmap. Try to set the bit.
* If that succeeds then check that nobody has allocated and then freed the
* block since we saw that is was not marked in b_committed_data. If it _was_
return ret;
}
-/*
+/**
+ * ext3_try_to_allocate()
+ * @sb: superblock
+ * @handle: handle to this transaction
+ * @group: given allocation block group
+ * @bitmap_bh: bufferhead holds the block bitmap
+ * @grp_goal: given target block within the group
+ * @count: target number of blocks to allocate
+ * @my_rsv: reservation window
+ *
+ * Attempt to allocate blocks within a give range. Set the range of allocation
+ * first, then find the first free bit(s) from the bitmap (within the range),
+ * and at last, allocate the blocks by claiming the found free bit as allocated.
+ *
+ * To set the range of this allocation:
+ * if there is a reservation window, only try to allocate block(s) from the
+ * file's own reservation window;
+ * Otherwise, the allocation range starts from the give goal block, ends at
+ * the block group's last block.
+ *
* If we failed to allocate the desired block then we may end up crossing to a
* new bitmap. In that case we must release write access to the old one via
* ext3_journal_release_buffer(), else we'll run out of credits.
}
start = grp_goal;
- if (!claim_block(sb_bgl_lock(EXT3_SB(sb), group), grp_goal, bitmap_bh)) {
+ if (!claim_block(sb_bgl_lock(EXT3_SB(sb), group),
+ grp_goal, bitmap_bh)) {
/*
* The block was allocated by another thread, or it was
* allocated and then freed by another thread
grp_goal++;
while (num < *count && grp_goal < end
&& ext3_test_allocatable(grp_goal, bitmap_bh)
- && claim_block(sb_bgl_lock(EXT3_SB(sb), group), grp_goal, bitmap_bh)) {
+ && claim_block(sb_bgl_lock(EXT3_SB(sb), group),
+ grp_goal, bitmap_bh)) {
num++;
grp_goal++;
}
}
/**
- * find_next_reservable_window():
+ * find_next_reservable_window():
* find a reservable space within the given range.
* It does not allocate the reservation window for now:
* alloc_new_reservation() will do the work later.
*
- * @search_head: the head of the searching list;
+ * @search_head: the head of the searching list;
* This is not necessarily the list head of the whole filesystem
*
* We have both head and start_block to assist the search
* but we will shift to the place where start_block is,
* then start from there, when looking for a reservable space.
*
- * @size: the target new reservation window size
+ * @size: the target new reservation window size
*
- * @group_first_block: the first block we consider to start
+ * @group_first_block: the first block we consider to start
* the real search from
*
- * @last_block:
+ * @last_block:
* the maximum block number that our goal reservable space
* could start from. This is normally the last block in this
* group. The search will end when we found the start of next
* This could handle the cross boundary reservation window
* request.
*
- * basically we search from the given range, rather than the whole
- * reservation double linked list, (start_block, last_block)
- * to find a free region that is of my size and has not
- * been reserved.
+ * basically we search from the given range, rather than the whole
+ * reservation double linked list, (start_block, last_block)
+ * to find a free region that is of my size and has not
+ * been reserved.
*
*/
static int find_next_reservable_window(
prev = rsv;
next = rb_next(&rsv->rsv_node);
- rsv = list_entry(next,struct ext3_reserve_window_node,rsv_node);
+ rsv = rb_entry(next,struct ext3_reserve_window_node,rsv_node);
/*
* Reached the last reservation, we can just append to the
/*
* Found a reserveable space big enough. We could
* have a reservation across the group boundary here
- */
+ */
break;
}
}
}
/**
- * alloc_new_reservation()--allocate a new reservation window
+ * alloc_new_reservation()--allocate a new reservation window
*
* To make a new reservation, we search part of the filesystem
* reservation list (the list that inside the group). We try to
spinlock_t *rsv_lock = &EXT3_SB(sb)->s_rsv_window_lock;
group_first_block = ext3_group_first_block_no(sb, group);
- group_end_block = group_first_block + EXT3_BLOCKS_PER_GROUP(sb) - 1;
+ group_end_block = group_first_block + (EXT3_BLOCKS_PER_GROUP(sb) - 1);
if (grp_goal < 0)
start_block = group_first_block;
if ((my_rsv->rsv_alloc_hit >
(my_rsv->rsv_end - my_rsv->rsv_start + 1) / 2)) {
/*
- * if we previously allocation hit ration is greater than half
- * we double the size of reservation window next time
- * otherwise keep the same
+ * if the previously allocation hit ratio is
+ * greater than 1/2, then we double the size of
+ * the reservation window the next time,
+ * otherwise we keep the same size window
*/
size = size * 2;
if (size > EXT3_MAX_RESERVE_BLOCKS)
* check if the first free block is within the
* free space we just reserved
*/
- if (start_block >= my_rsv->rsv_start && start_block < my_rsv->rsv_end)
+ if (start_block >= my_rsv->rsv_start && start_block <= my_rsv->rsv_end)
return 0; /* success */
/*
* if the first free bit we found is out of the reservable space
goto retry;
}
+/**
+ * try_to_extend_reservation()
+ * @my_rsv: given reservation window
+ * @sb: super block
+ * @size: the delta to extend
+ *
+ * Attempt to expand the reservation window large enough to have
+ * required number of free blocks
+ *
+ * Since ext3_try_to_allocate() will always allocate blocks within
+ * the reservation window range, if the window size is too small,
+ * multiple blocks allocation has to stop at the end of the reservation
+ * window. To make this more efficient, given the total number of
+ * blocks needed and the current size of the window, we try to
+ * expand the reservation window size if necessary on a best-effort
+ * basis before ext3_new_blocks() tries to allocate blocks,
+ */
static void try_to_extend_reservation(struct ext3_reserve_window_node *my_rsv,
struct super_block *sb, int size)
{
if (!next)
my_rsv->rsv_end += size;
else {
- next_rsv = list_entry(next, struct ext3_reserve_window_node, rsv_node);
+ next_rsv = rb_entry(next, struct ext3_reserve_window_node, rsv_node);
if ((next_rsv->rsv_start - my_rsv->rsv_end - 1) >= size)
my_rsv->rsv_end += size;
spin_unlock(rsv_lock);
}
-/*
+/**
+ * ext3_try_to_allocate_with_rsv()
+ * @sb: superblock
+ * @handle: handle to this transaction
+ * @group: given allocation block group
+ * @bitmap_bh: bufferhead holds the block bitmap
+ * @grp_goal: given target block within the group
+ * @count: target number of blocks to allocate
+ * @my_rsv: reservation window
+ * @errp: pointer to store the error code
+ *
* This is the main function used to allocate a new block and its reservation
* window.
*
* reservation), and there are lots of free blocks, but they are all
* being reserved.
*
- * We use a sorted double linked list for the per-filesystem reservation list.
- * The insert, remove and find a free space(non-reserved) operations for the
- * sorted double linked list should be fast.
+ * We use a red-black tree for the per-filesystem reservation list.
*
*/
static ext3_grpblk_t
struct ext3_reserve_window_node * my_rsv,
unsigned long *count, int *errp)
{
- ext3_fsblk_t group_first_block;
+ ext3_fsblk_t group_first_block, group_last_block;
ext3_grpblk_t ret = 0;
int fatal;
unsigned long num = *count;
}
/*
* grp_goal is a group relative block number (if there is a goal)
- * 0 < grp_goal < EXT3_BLOCKS_PER_GROUP(sb)
+ * 0 <= grp_goal < EXT3_BLOCKS_PER_GROUP(sb)
* first block is a filesystem wide block number
* first block is the block number of the first block in this group
*/
group_first_block = ext3_group_first_block_no(sb, group);
+ group_last_block = group_first_block + (EXT3_BLOCKS_PER_GROUP(sb) - 1);
/*
* Basically we will allocate a new block from inode's reservation
*/
while (1) {
if (rsv_is_empty(&my_rsv->rsv_window) || (ret < 0) ||
- !goal_in_my_reservation(&my_rsv->rsv_window, grp_goal, group, sb)) {
+ !goal_in_my_reservation(&my_rsv->rsv_window,
+ grp_goal, group, sb)) {
if (my_rsv->rsv_goal_size < *count)
my_rsv->rsv_goal_size = *count;
ret = alloc_new_reservation(my_rsv, grp_goal, sb,
if (ret < 0)
break; /* failed */
- if (!goal_in_my_reservation(&my_rsv->rsv_window, grp_goal, group, sb))
+ if (!goal_in_my_reservation(&my_rsv->rsv_window,
+ grp_goal, group, sb))
grp_goal = -1;
- } else if (grp_goal > 0 && (my_rsv->rsv_end-grp_goal+1) < *count)
- try_to_extend_reservation(my_rsv, sb,
- *count-my_rsv->rsv_end + grp_goal - 1);
+ } else if (grp_goal >= 0) {
+ int curr = my_rsv->rsv_end -
+ (grp_goal + group_first_block) + 1;
+
+ if (curr < *count)
+ try_to_extend_reservation(my_rsv, sb,
+ *count - curr);
+ }
- if ((my_rsv->rsv_start >= group_first_block + EXT3_BLOCKS_PER_GROUP(sb))
- || (my_rsv->rsv_end < group_first_block))
+ if ((my_rsv->rsv_start > group_last_block) ||
+ (my_rsv->rsv_end < group_first_block)) {
+ rsv_window_dump(&EXT3_SB(sb)->s_rsv_window_root, 1);
BUG();
- ret = ext3_try_to_allocate(sb, handle, group, bitmap_bh, grp_goal,
- &num, &my_rsv->rsv_window);
+ }
+ ret = ext3_try_to_allocate(sb, handle, group, bitmap_bh,
+ grp_goal, &num, &my_rsv->rsv_window);
if (ret >= 0) {
my_rsv->rsv_alloc_hit += num;
*count = num;
return ret;
}
+/**
+ * ext3_has_free_blocks()
+ * @sbi: in-core super block structure.
+ *
+ * Check if filesystem has at least 1 free block available for allocation.
+ */
static int ext3_has_free_blocks(struct ext3_sb_info *sbi)
{
ext3_fsblk_t free_blocks, root_blocks;
return 1;
}
-/*
+/**
+ * ext3_should_retry_alloc()
+ * @sb: super block
+ * @retries number of attemps has been made
+ *
* ext3_should_retry_alloc() is called when ENOSPC is returned, and if
* it is profitable to retry the operation, this function will wait
* for the current or commiting transaction to complete, and then
* return TRUE.
+ *
+ * if the total number of retries exceed three times, return FALSE.
*/
int ext3_should_retry_alloc(struct super_block *sb, int *retries)
{
return journal_force_commit_nested(EXT3_SB(sb)->s_journal);
}
-/*
- * ext3_new_block uses a goal block to assist allocation. If the goal is
- * free, or there is a free block within 32 blocks of the goal, that block
- * is allocated. Otherwise a forward search is made for a free block; within
- * each block group the search first looks for an entire free byte in the block
- * bitmap, and then for any free bit if that fails.
- * This function also updates quota and i_blocks field.
+/**
+ * ext3_new_blocks() -- core block(s) allocation function
+ * @handle: handle to this transaction
+ * @inode: file inode
+ * @goal: given target block(filesystem wide)
+ * @count: target number of blocks to allocate
+ * @errp: error code
+ *
+ * ext3_new_blocks uses a goal block to assist allocation. It tries to
+ * allocate block(s) from the block group contains the goal block first. If that
+ * fails, it will try to allocate block(s) from other block groups without
+ * any specific goal block.
+ *
*/
ext3_fsblk_t ext3_new_blocks(handle_t *handle, struct inode *inode,
ext3_fsblk_t goal, unsigned long *count, int *errp)
goal = le32_to_cpu(es->s_first_data_block);
group_no = (goal - le32_to_cpu(es->s_first_data_block)) /
EXT3_BLOCKS_PER_GROUP(sb);
+ goal_group = group_no;
+retry_alloc:
gdp = ext3_get_group_desc(sb, group_no, &gdp_bh);
if (!gdp)
goto io_error;
- goal_group = group_no;
-retry:
free_blocks = le16_to_cpu(gdp->bg_free_blocks_count);
/*
* if there is not enough free blocks to make a new resevation
smp_rmb();
/*
- * Now search the rest of the groups. We assume that
- * i and gdp correctly point to the last group visited.
+ * Now search the rest of the groups. We assume that
+ * group_no and gdp correctly point to the last group visited.
*/
for (bgi = 0; bgi < ngroups; bgi++) {
group_no++;
if (group_no >= ngroups)
group_no = 0;
gdp = ext3_get_group_desc(sb, group_no, &gdp_bh);
- if (!gdp) {
- *errp = -EIO;
- goto out;
- }
+ if (!gdp)
+ goto io_error;
free_blocks = le16_to_cpu(gdp->bg_free_blocks_count);
/*
* skip this group if the number of
*/
if (my_rsv) {
my_rsv = NULL;
+ windowsz = 0;
group_no = goal_group;
- goto retry;
+ goto retry_alloc;
}
/* No space left on the device */
*errp = -ENOSPC;
in_range(ret_block, le32_to_cpu(gdp->bg_inode_table),
EXT3_SB(sb)->s_itb_per_group) ||
in_range(ret_block + num - 1, le32_to_cpu(gdp->bg_inode_table),
- EXT3_SB(sb)->s_itb_per_group))
+ EXT3_SB(sb)->s_itb_per_group)) {
ext3_error(sb, "ext3_new_block",
"Allocating block in system zone - "
"blocks from "E3FSBLK", length %lu",
ret_block, num);
+ /*
+ * claim_block() marked the blocks we allocated as in use. So we
+ * may want to selectively mark some of the blocks as free.
+ */
+ goto retry_alloc;
+ }
performed_allocation = 1;
if (ext3_test_bit(grp_alloc_blk+i,
bh2jh(bitmap_bh)->b_committed_data)) {
printk("%s: block was unexpectedly set in "
- "b_committed_data\n", __FUNCTION__);
+ "b_committed_data\n", __func__);
}
}
}
ret_block, goal_hits, goal_attempts);
spin_lock(sb_bgl_lock(sbi, group_no));
- gdp->bg_free_blocks_count =
- cpu_to_le16(le16_to_cpu(gdp->bg_free_blocks_count) - num);
+ le16_add_cpu(&gdp->bg_free_blocks_count, -num);
spin_unlock(sb_bgl_lock(sbi, group_no));
- percpu_counter_mod(&sbi->s_freeblocks_counter, -num);
+ percpu_counter_sub(&sbi->s_freeblocks_counter, num);
BUFFER_TRACE(gdp_bh, "journal_dirty_metadata for group descriptor");
err = ext3_journal_dirty_metadata(handle, gdp_bh);
return ext3_new_blocks(handle, inode, goal, &count, errp);
}
+/**
+ * ext3_count_free_blocks() -- count filesystem free blocks
+ * @sb: superblock
+ *
+ * Adds up the number of free blocks from each block group.
+ */
ext3_fsblk_t ext3_count_free_blocks(struct super_block *sb)
{
ext3_fsblk_t desc_count;
#endif
}
-static inline int
-block_in_use(ext3_fsblk_t block, struct super_block *sb, unsigned char *map)
-{
- return ext3_test_bit ((block -
- le32_to_cpu(EXT3_SB(sb)->s_es->s_first_data_block)) %
- EXT3_BLOCKS_PER_GROUP(sb), map);
-}
-
static inline int test_root(int a, int b)
{
int num = b;
static unsigned long ext3_bg_num_gdb_nometa(struct super_block *sb, int group)
{
- if (EXT3_HAS_RO_COMPAT_FEATURE(sb,
- EXT3_FEATURE_RO_COMPAT_SPARSE_SUPER) &&
- !ext3_group_sparse(group))
- return 0;
- return EXT3_SB(sb)->s_gdb_count;
+ return ext3_bg_has_super(sb, group) ? EXT3_SB(sb)->s_gdb_count : 0;
}
/**