]> www.pilppa.org Git - linux-2.6-omap-h63xx.git/blobdiff - fs/ext4/extents.c
ext4: move headers out of include/linux
[linux-2.6-omap-h63xx.git] / fs / ext4 / extents.c
index 85287742f2ae487562be3a8ac39e40774a5de9d3..a472bc046363bcf2563d688367da7f196c010362 100644 (file)
@@ -32,7 +32,6 @@
 #include <linux/module.h>
 #include <linux/fs.h>
 #include <linux/time.h>
-#include <linux/ext4_jbd2.h>
 #include <linux/jbd2.h>
 #include <linux/highuid.h>
 #include <linux/pagemap.h>
@@ -40,8 +39,9 @@
 #include <linux/string.h>
 #include <linux/slab.h>
 #include <linux/falloc.h>
-#include <linux/ext4_fs_extents.h>
 #include <asm/uaccess.h>
+#include "ext4_jbd2.h"
+#include "ext4_extents.h"
 
 
 /*
@@ -61,7 +61,7 @@ static ext4_fsblk_t ext_pblock(struct ext4_extent *ex)
  * idx_pblock:
  * combine low and high parts of a leaf physical block number into ext4_fsblk_t
  */
-static ext4_fsblk_t idx_pblock(struct ext4_extent_idx *ix)
+ext4_fsblk_t idx_pblock(struct ext4_extent_idx *ix)
 {
        ext4_fsblk_t block;
 
@@ -75,7 +75,7 @@ static ext4_fsblk_t idx_pblock(struct ext4_extent_idx *ix)
  * stores a large physical block number into an extent struct,
  * breaking it into parts
  */
-static void ext4_ext_store_pblock(struct ext4_extent *ex, ext4_fsblk_t pb)
+void ext4_ext_store_pblock(struct ext4_extent *ex, ext4_fsblk_t pb)
 {
        ex->ee_start_lo = cpu_to_le32((unsigned long) (pb & 0xffffffff));
        ex->ee_start_hi = cpu_to_le16((unsigned long) ((pb >> 31) >> 1) & 0xffff);
@@ -144,10 +144,11 @@ static int ext4_ext_dirty(handle_t *handle, struct inode *inode,
 
 static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
                              struct ext4_ext_path *path,
-                             ext4_fsblk_t block)
+                             ext4_lblk_t block)
 {
        struct ext4_inode_info *ei = EXT4_I(inode);
        ext4_fsblk_t bg_start;
+       ext4_fsblk_t last_block;
        ext4_grpblk_t colour;
        int depth;
 
@@ -169,8 +170,13 @@ static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode,
        /* OK. use inode's group */
        bg_start = (ei->i_block_group * EXT4_BLOCKS_PER_GROUP(inode->i_sb)) +
                le32_to_cpu(EXT4_SB(inode->i_sb)->s_es->s_first_data_block);
-       colour = (current->pid % 16) *
+       last_block = ext4_blocks_count(EXT4_SB(inode->i_sb)->s_es) - 1;
+
+       if (bg_start + EXT4_BLOCKS_PER_GROUP(inode->i_sb) <= last_block)
+               colour = (current->pid % 16) *
                        (EXT4_BLOCKS_PER_GROUP(inode->i_sb) / 16);
+       else
+               colour = (current->pid % 16) * ((last_block - bg_start) / 16);
        return bg_start + colour + block;
 }
 
@@ -302,7 +308,7 @@ corrupted:
 }
 
 #define ext4_ext_check_header(inode, eh, depth)        \
-       __ext4_ext_check_header(__FUNCTION__, inode, eh, depth)
+       __ext4_ext_check_header(__func__, inode, eh, depth)
 
 #ifdef EXT_DEBUG
 static void ext4_ext_show_path(struct inode *inode, struct ext4_ext_path *path)
@@ -349,7 +355,7 @@ static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path)
 #define ext4_ext_show_leaf(inode,path)
 #endif
 
-static void ext4_ext_drop_refs(struct ext4_ext_path *path)
+void ext4_ext_drop_refs(struct ext4_ext_path *path)
 {
        int depth = path->p_depth;
        int i;
@@ -367,13 +373,14 @@ static void ext4_ext_drop_refs(struct ext4_ext_path *path)
  * the header must be checked before calling this
  */
 static void
-ext4_ext_binsearch_idx(struct inode *inode, struct ext4_ext_path *path, int block)
+ext4_ext_binsearch_idx(struct inode *inode,
+                       struct ext4_ext_path *path, ext4_lblk_t block)
 {
        struct ext4_extent_header *eh = path->p_hdr;
        struct ext4_extent_idx *r, *l, *m;
 
 
-       ext_debug("binsearch for %d(idx):  ", block);
+       ext_debug("binsearch for %u(idx):  ", block);
 
        l = EXT_FIRST_INDEX(eh) + 1;
        r = EXT_LAST_INDEX(eh);
@@ -425,7 +432,8 @@ ext4_ext_binsearch_idx(struct inode *inode, struct ext4_ext_path *path, int bloc
  * the header must be checked before calling this
  */
 static void
-ext4_ext_binsearch(struct inode *inode, struct ext4_ext_path *path, int block)
+ext4_ext_binsearch(struct inode *inode,
+               struct ext4_ext_path *path, ext4_lblk_t block)
 {
        struct ext4_extent_header *eh = path->p_hdr;
        struct ext4_extent *r, *l, *m;
@@ -438,7 +446,7 @@ ext4_ext_binsearch(struct inode *inode, struct ext4_ext_path *path, int block)
                return;
        }
 
-       ext_debug("binsearch for %d:  ", block);
+       ext_debug("binsearch for %u:  ", block);
 
        l = EXT_FIRST_EXTENT(eh) + 1;
        r = EXT_LAST_EXTENT(eh);
@@ -494,7 +502,8 @@ int ext4_ext_tree_init(handle_t *handle, struct inode *inode)
 }
 
 struct ext4_ext_path *
-ext4_ext_find_extent(struct inode *inode, int block, struct ext4_ext_path *path)
+ext4_ext_find_extent(struct inode *inode, ext4_lblk_t block,
+                                       struct ext4_ext_path *path)
 {
        struct ext4_extent_header *eh;
        struct buffer_head *bh;
@@ -605,7 +614,7 @@ static int ext4_ext_insert_index(handle_t *handle, struct inode *inode,
 
        ix->ei_block = cpu_to_le32(logical);
        ext4_idx_store_pblock(ix, ptr);
-       curp->p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(curp->p_hdr->eh_entries)+1);
+       le16_add_cpu(&curp->p_hdr->eh_entries, 1);
 
        BUG_ON(le16_to_cpu(curp->p_hdr->eh_entries)
                             > le16_to_cpu(curp->p_hdr->eh_max));
@@ -727,7 +736,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
        }
        if (m) {
                memmove(ex, path[depth].p_ext-m, sizeof(struct ext4_extent)*m);
-               neh->eh_entries = cpu_to_le16(le16_to_cpu(neh->eh_entries)+m);
+               le16_add_cpu(&neh->eh_entries, m);
        }
 
        set_buffer_uptodate(bh);
@@ -744,8 +753,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
                err = ext4_ext_get_access(handle, inode, path + depth);
                if (err)
                        goto cleanup;
-               path[depth].p_hdr->eh_entries =
-                    cpu_to_le16(le16_to_cpu(path[depth].p_hdr->eh_entries)-m);
+               le16_add_cpu(&path[depth].p_hdr->eh_entries, -m);
                err = ext4_ext_dirty(handle, inode, path + depth);
                if (err)
                        goto cleanup;
@@ -763,7 +771,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
        while (k--) {
                oldblock = newblock;
                newblock = ablocks[--a];
-               bh = sb_getblk(inode->i_sb, (ext4_fsblk_t)newblock);
+               bh = sb_getblk(inode->i_sb, newblock);
                if (!bh) {
                        err = -EIO;
                        goto cleanup;
@@ -783,9 +791,8 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
                fidx->ei_block = border;
                ext4_idx_store_pblock(fidx, oldblock);
 
-               ext_debug("int.index at %d (block %llu): %lu -> %llu\n", i,
-                               newblock, (unsigned long) le32_to_cpu(border),
-                               oldblock);
+               ext_debug("int.index at %d (block %llu): %u -> %llu\n",
+                               i, newblock, le32_to_cpu(border), oldblock);
                /* copy indexes */
                m = 0;
                path[i].p_idx++;
@@ -809,8 +816,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
                if (m) {
                        memmove(++fidx, path[i].p_idx - m,
                                sizeof(struct ext4_extent_idx) * m);
-                       neh->eh_entries =
-                               cpu_to_le16(le16_to_cpu(neh->eh_entries) + m);
+                       le16_add_cpu(&neh->eh_entries, m);
                }
                set_buffer_uptodate(bh);
                unlock_buffer(bh);
@@ -826,7 +832,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode,
                        err = ext4_ext_get_access(handle, inode, path + i);
                        if (err)
                                goto cleanup;
-                       path[i].p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(path[i].p_hdr->eh_entries)-m);
+                       le16_add_cpu(&path[i].p_hdr->eh_entries, -m);
                        err = ext4_ext_dirty(handle, inode, path + i);
                        if (err)
                                goto cleanup;
@@ -851,7 +857,7 @@ cleanup:
                for (i = 0; i < depth; i++) {
                        if (!ablocks[i])
                                continue;
-                       ext4_free_blocks(handle, inode, ablocks[i], 1);
+                       ext4_free_blocks(handle, inode, ablocks[i], 1, 1);
                }
        }
        kfree(ablocks);
@@ -979,8 +985,8 @@ repeat:
                /* refill path */
                ext4_ext_drop_refs(path);
                path = ext4_ext_find_extent(inode,
-                                           le32_to_cpu(newext->ee_block),
-                                           path);
+                                   (ext4_lblk_t)le32_to_cpu(newext->ee_block),
+                                   path);
                if (IS_ERR(path))
                        err = PTR_ERR(path);
        } else {
@@ -992,8 +998,8 @@ repeat:
                /* refill path */
                ext4_ext_drop_refs(path);
                path = ext4_ext_find_extent(inode,
-                                           le32_to_cpu(newext->ee_block),
-                                           path);
+                                  (ext4_lblk_t)le32_to_cpu(newext->ee_block),
+                                   path);
                if (IS_ERR(path)) {
                        err = PTR_ERR(path);
                        goto out;
@@ -1014,6 +1020,150 @@ out:
        return err;
 }
 
+/*
+ * search the closest allocated block to the left for *logical
+ * and returns it at @logical + it's physical address at @phys
+ * if *logical is the smallest allocated block, the function
+ * returns 0 at @phys
+ * return value contains 0 (success) or error code
+ */
+int
+ext4_ext_search_left(struct inode *inode, struct ext4_ext_path *path,
+                       ext4_lblk_t *logical, ext4_fsblk_t *phys)
+{
+       struct ext4_extent_idx *ix;
+       struct ext4_extent *ex;
+       int depth, ee_len;
+
+       BUG_ON(path == NULL);
+       depth = path->p_depth;
+       *phys = 0;
+
+       if (depth == 0 && path->p_ext == NULL)
+               return 0;
+
+       /* usually extent in the path covers blocks smaller
+        * then *logical, but it can be that extent is the
+        * first one in the file */
+
+       ex = path[depth].p_ext;
+       ee_len = ext4_ext_get_actual_len(ex);
+       if (*logical < le32_to_cpu(ex->ee_block)) {
+               BUG_ON(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex);
+               while (--depth >= 0) {
+                       ix = path[depth].p_idx;
+                       BUG_ON(ix != EXT_FIRST_INDEX(path[depth].p_hdr));
+               }
+               return 0;
+       }
+
+       BUG_ON(*logical < (le32_to_cpu(ex->ee_block) + ee_len));
+
+       *logical = le32_to_cpu(ex->ee_block) + ee_len - 1;
+       *phys = ext_pblock(ex) + ee_len - 1;
+       return 0;
+}
+
+/*
+ * search the closest allocated block to the right for *logical
+ * and returns it at @logical + it's physical address at @phys
+ * if *logical is the smallest allocated block, the function
+ * returns 0 at @phys
+ * return value contains 0 (success) or error code
+ */
+int
+ext4_ext_search_right(struct inode *inode, struct ext4_ext_path *path,
+                       ext4_lblk_t *logical, ext4_fsblk_t *phys)
+{
+       struct buffer_head *bh = NULL;
+       struct ext4_extent_header *eh;
+       struct ext4_extent_idx *ix;
+       struct ext4_extent *ex;
+       ext4_fsblk_t block;
+       int depth, ee_len;
+
+       BUG_ON(path == NULL);
+       depth = path->p_depth;
+       *phys = 0;
+
+       if (depth == 0 && path->p_ext == NULL)
+               return 0;
+
+       /* usually extent in the path covers blocks smaller
+        * then *logical, but it can be that extent is the
+        * first one in the file */
+
+       ex = path[depth].p_ext;
+       ee_len = ext4_ext_get_actual_len(ex);
+       if (*logical < le32_to_cpu(ex->ee_block)) {
+               BUG_ON(EXT_FIRST_EXTENT(path[depth].p_hdr) != ex);
+               while (--depth >= 0) {
+                       ix = path[depth].p_idx;
+                       BUG_ON(ix != EXT_FIRST_INDEX(path[depth].p_hdr));
+               }
+               *logical = le32_to_cpu(ex->ee_block);
+               *phys = ext_pblock(ex);
+               return 0;
+       }
+
+       BUG_ON(*logical < (le32_to_cpu(ex->ee_block) + ee_len));
+
+       if (ex != EXT_LAST_EXTENT(path[depth].p_hdr)) {
+               /* next allocated block in this leaf */
+               ex++;
+               *logical = le32_to_cpu(ex->ee_block);
+               *phys = ext_pblock(ex);
+               return 0;
+       }
+
+       /* go up and search for index to the right */
+       while (--depth >= 0) {
+               ix = path[depth].p_idx;
+               if (ix != EXT_LAST_INDEX(path[depth].p_hdr))
+                       break;
+       }
+
+       if (depth < 0) {
+               /* we've gone up to the root and
+                * found no index to the right */
+               return 0;
+       }
+
+       /* we've found index to the right, let's
+        * follow it and find the closest allocated
+        * block to the right */
+       ix++;
+       block = idx_pblock(ix);
+       while (++depth < path->p_depth) {
+               bh = sb_bread(inode->i_sb, block);
+               if (bh == NULL)
+                       return -EIO;
+               eh = ext_block_hdr(bh);
+               if (ext4_ext_check_header(inode, eh, depth)) {
+                       put_bh(bh);
+                       return -EIO;
+               }
+               ix = EXT_FIRST_INDEX(eh);
+               block = idx_pblock(ix);
+               put_bh(bh);
+       }
+
+       bh = sb_bread(inode->i_sb, block);
+       if (bh == NULL)
+               return -EIO;
+       eh = ext_block_hdr(bh);
+       if (ext4_ext_check_header(inode, eh, path->p_depth - depth)) {
+               put_bh(bh);
+               return -EIO;
+       }
+       ex = EXT_FIRST_EXTENT(eh);
+       *logical = le32_to_cpu(ex->ee_block);
+       *phys = ext_pblock(ex);
+       put_bh(bh);
+       return 0;
+
+}
+
 /*
  * ext4_ext_next_allocated_block:
  * returns allocated block in subsequent extent or EXT_MAX_BLOCK.
@@ -1021,7 +1171,7 @@ out:
  * allocated block. Thus, index entries have to be consistent
  * with leaves.
  */
-static unsigned long
+static ext4_lblk_t
 ext4_ext_next_allocated_block(struct ext4_ext_path *path)
 {
        int depth;
@@ -1054,7 +1204,7 @@ ext4_ext_next_allocated_block(struct ext4_ext_path *path)
  * ext4_ext_next_leaf_block:
  * returns first allocated block from next leaf or EXT_MAX_BLOCK
  */
-static unsigned ext4_ext_next_leaf_block(struct inode *inode,
+static ext4_lblk_t ext4_ext_next_leaf_block(struct inode *inode,
                                        struct ext4_ext_path *path)
 {
        int depth;
@@ -1072,7 +1222,8 @@ static unsigned ext4_ext_next_leaf_block(struct inode *inode,
        while (depth >= 0) {
                if (path[depth].p_idx !=
                                EXT_LAST_INDEX(path[depth].p_hdr))
-                 return le32_to_cpu(path[depth].p_idx[1].ei_block);
+                       return (ext4_lblk_t)
+                               le32_to_cpu(path[depth].p_idx[1].ei_block);
                depth--;
        }
 
@@ -1085,7 +1236,7 @@ static unsigned ext4_ext_next_leaf_block(struct inode *inode,
  * then we have to correct all indexes above.
  * TODO: do we need to correct tree in all cases?
  */
-int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
+static int ext4_ext_correct_indexes(handle_t *handle, struct inode *inode,
                                struct ext4_ext_path *path)
 {
        struct ext4_extent_header *eh;
@@ -1171,7 +1322,7 @@ ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1,
        if (ext1_ee_len + ext2_ee_len > max_len)
                return 0;
 #ifdef AGGRESSIVE_TEST
-       if (le16_to_cpu(ex1->ee_len) >= 4)
+       if (ext1_ee_len >= 4)
                return 0;
 #endif
 
@@ -1216,7 +1367,7 @@ int ext4_ext_try_to_merge(struct inode *inode,
                                * sizeof(struct ext4_extent);
                        memmove(ex + 1, ex + 2, len);
                }
-               eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries) - 1);
+               le16_add_cpu(&eh->eh_entries, -1);
                merge_done = 1;
                WARN_ON(eh->eh_entries == 0);
                if (!eh->eh_entries)
@@ -1239,7 +1390,7 @@ unsigned int ext4_ext_check_overlap(struct inode *inode,
                                    struct ext4_extent *newext,
                                    struct ext4_ext_path *path)
 {
-       unsigned long b1, b2;
+       ext4_lblk_t b1, b2;
        unsigned int depth, len1;
        unsigned int ret = 0;
 
@@ -1260,7 +1411,7 @@ unsigned int ext4_ext_check_overlap(struct inode *inode,
                        goto out;
        }
 
-       /* check for wrap through zero */
+       /* check for wrap through zero on extent logical start block*/
        if (b1 + len1 < b1) {
                len1 = EXT_MAX_BLOCK - b1;
                newext->ee_len = cpu_to_le16(len1);
@@ -1290,7 +1441,8 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode,
        struct ext4_extent *ex, *fex;
        struct ext4_extent *nearex; /* nearest extent */
        struct ext4_ext_path *npath = NULL;
-       int depth, len, err, next;
+       int depth, len, err;
+       ext4_lblk_t next;
        unsigned uninitialized = 0;
 
        BUG_ON(ext4_ext_get_actual_len(newext) == 0);
@@ -1406,7 +1558,7 @@ has_space:
                path[depth].p_ext = nearex;
        }
 
-       eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries)+1);
+       le16_add_cpu(&eh->eh_entries, 1);
        nearex = path[depth].p_ext;
        nearex->ee_block = newext->ee_block;
        ext4_ext_store_pblock(nearex, ext_pblock(newext));
@@ -1435,114 +1587,8 @@ cleanup:
        return err;
 }
 
-int ext4_ext_walk_space(struct inode *inode, unsigned long block,
-                       unsigned long num, ext_prepare_callback func,
-                       void *cbdata)
-{
-       struct ext4_ext_path *path = NULL;
-       struct ext4_ext_cache cbex;
-       struct ext4_extent *ex;
-       unsigned long next, start = 0, end = 0;
-       unsigned long last = block + num;
-       int depth, exists, err = 0;
-
-       BUG_ON(func == NULL);
-       BUG_ON(inode == NULL);
-
-       while (block < last && block != EXT_MAX_BLOCK) {
-               num = last - block;
-               /* find extent for this block */
-               path = ext4_ext_find_extent(inode, block, path);
-               if (IS_ERR(path)) {
-                       err = PTR_ERR(path);
-                       path = NULL;
-                       break;
-               }
-
-               depth = ext_depth(inode);
-               BUG_ON(path[depth].p_hdr == NULL);
-               ex = path[depth].p_ext;
-               next = ext4_ext_next_allocated_block(path);
-
-               exists = 0;
-               if (!ex) {
-                       /* there is no extent yet, so try to allocate
-                        * all requested space */
-                       start = block;
-                       end = block + num;
-               } else if (le32_to_cpu(ex->ee_block) > block) {
-                       /* need to allocate space before found extent */
-                       start = block;
-                       end = le32_to_cpu(ex->ee_block);
-                       if (block + num < end)
-                               end = block + num;
-               } else if (block >= le32_to_cpu(ex->ee_block)
-                                       + ext4_ext_get_actual_len(ex)) {
-                       /* need to allocate space after found extent */
-                       start = block;
-                       end = block + num;
-                       if (end >= next)
-                               end = next;
-               } else if (block >= le32_to_cpu(ex->ee_block)) {
-                       /*
-                        * some part of requested space is covered
-                        * by found extent
-                        */
-                       start = block;
-                       end = le32_to_cpu(ex->ee_block)
-                               + ext4_ext_get_actual_len(ex);
-                       if (block + num < end)
-                               end = block + num;
-                       exists = 1;
-               } else {
-                       BUG();
-               }
-               BUG_ON(end <= start);
-
-               if (!exists) {
-                       cbex.ec_block = start;
-                       cbex.ec_len = end - start;
-                       cbex.ec_start = 0;
-                       cbex.ec_type = EXT4_EXT_CACHE_GAP;
-               } else {
-                       cbex.ec_block = le32_to_cpu(ex->ee_block);
-                       cbex.ec_len = ext4_ext_get_actual_len(ex);
-                       cbex.ec_start = ext_pblock(ex);
-                       cbex.ec_type = EXT4_EXT_CACHE_EXTENT;
-               }
-
-               BUG_ON(cbex.ec_len == 0);
-               err = func(inode, path, &cbex, cbdata);
-               ext4_ext_drop_refs(path);
-
-               if (err < 0)
-                       break;
-               if (err == EXT_REPEAT)
-                       continue;
-               else if (err == EXT_BREAK) {
-                       err = 0;
-                       break;
-               }
-
-               if (ext_depth(inode) != depth) {
-                       /* depth was changed. we have to realloc path */
-                       kfree(path);
-                       path = NULL;
-               }
-
-               block = cbex.ec_block + cbex.ec_len;
-       }
-
-       if (path) {
-               ext4_ext_drop_refs(path);
-               kfree(path);
-       }
-
-       return err;
-}
-
 static void
-ext4_ext_put_in_cache(struct inode *inode, __u32 block,
+ext4_ext_put_in_cache(struct inode *inode, ext4_lblk_t block,
                        __u32 len, ext4_fsblk_t start, int type)
 {
        struct ext4_ext_cache *cex;
@@ -1561,10 +1607,11 @@ ext4_ext_put_in_cache(struct inode *inode, __u32 block,
  */
 static void
 ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
-                               unsigned long block)
+                               ext4_lblk_t block)
 {
        int depth = ext_depth(inode);
-       unsigned long lblock, len;
+       unsigned long len;
+       ext4_lblk_t lblock;
        struct ext4_extent *ex;
 
        ex = path[depth].p_ext;
@@ -1576,32 +1623,34 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path,
        } else if (block < le32_to_cpu(ex->ee_block)) {
                lblock = block;
                len = le32_to_cpu(ex->ee_block) - block;
-               ext_debug("cache gap(before): %lu [%lu:%lu]",
-                               (unsigned long) block,
-                               (unsigned long) le32_to_cpu(ex->ee_block),
-                               (unsigned long) ext4_ext_get_actual_len(ex));
+               ext_debug("cache gap(before): %u [%u:%u]",
+                               block,
+                               le32_to_cpu(ex->ee_block),
+                                ext4_ext_get_actual_len(ex));
        } else if (block >= le32_to_cpu(ex->ee_block)
                        + ext4_ext_get_actual_len(ex)) {
+               ext4_lblk_t next;
                lblock = le32_to_cpu(ex->ee_block)
                        + ext4_ext_get_actual_len(ex);
-               len = ext4_ext_next_allocated_block(path);
-               ext_debug("cache gap(after): [%lu:%lu] %lu",
-                               (unsigned long) le32_to_cpu(ex->ee_block),
-                               (unsigned long) ext4_ext_get_actual_len(ex),
-                               (unsigned long) block);
-               BUG_ON(len == lblock);
-               len = len - lblock;
+
+               next = ext4_ext_next_allocated_block(path);
+               ext_debug("cache gap(after): [%u:%u] %u",
+                               le32_to_cpu(ex->ee_block),
+                               ext4_ext_get_actual_len(ex),
+                               block);
+               BUG_ON(next == lblock);
+               len = next - lblock;
        } else {
                lblock = len = 0;
                BUG();
        }
 
-       ext_debug(" -> %lu:%lu\n", (unsigned long) lblock, len);
+       ext_debug(" -> %u:%lu\n", lblock, len);
        ext4_ext_put_in_cache(inode, lblock, len, 0, EXT4_EXT_CACHE_GAP);
 }
 
 static int
-ext4_ext_in_cache(struct inode *inode, unsigned long block,
+ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block,
                        struct ext4_extent *ex)
 {
        struct ext4_ext_cache *cex;
@@ -1618,11 +1667,9 @@ ext4_ext_in_cache(struct inode *inode, unsigned long block,
                ex->ee_block = cpu_to_le32(cex->ec_block);
                ext4_ext_store_pblock(ex, cex->ec_start);
                ex->ee_len = cpu_to_le16(cex->ec_len);
-               ext_debug("%lu cached by %lu:%lu:%llu\n",
-                               (unsigned long) block,
-                               (unsigned long) cex->ec_block,
-                               (unsigned long) cex->ec_len,
-                               cex->ec_start);
+               ext_debug("%u cached by %u:%u:%llu\n",
+                               block,
+                               cex->ec_block, cex->ec_len, cex->ec_start);
                return cex->ec_type;
        }
 
@@ -1636,7 +1683,7 @@ ext4_ext_in_cache(struct inode *inode, unsigned long block,
  * It's used in truncate case only, thus all requests are for
  * last index in the block only.
  */
-int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
+static int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
                        struct ext4_ext_path *path)
 {
        struct buffer_head *bh;
@@ -1650,14 +1697,14 @@ int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
        err = ext4_ext_get_access(handle, inode, path);
        if (err)
                return err;
-       path->p_hdr->eh_entries = cpu_to_le16(le16_to_cpu(path->p_hdr->eh_entries)-1);
+       le16_add_cpu(&path->p_hdr->eh_entries, -1);
        err = ext4_ext_dirty(handle, inode, path);
        if (err)
                return err;
        ext_debug("index is empty, remove it, free block %llu\n", leaf);
        bh = sb_find_get_block(inode->i_sb, leaf);
        ext4_forget(handle, 1, inode, bh, leaf);
-       ext4_free_blocks(handle, inode, leaf, 1);
+       ext4_free_blocks(handle, inode, leaf, 1, 1);
        return err;
 }
 
@@ -1666,7 +1713,7 @@ int ext4_ext_rm_idx(handle_t *handle, struct inode *inode,
  * This routine returns max. credits that the extent tree can consume.
  * It should be OK for low-performance paths like ->writepage()
  * To allow many writing processes to fit into a single transaction,
- * the caller should calculate credits under truncate_mutex and
+ * the caller should calculate credits under i_data_sem and
  * pass the actual path.
  */
 int ext4_ext_calc_credits_for_insert(struct inode *inode,
@@ -1714,12 +1761,14 @@ int ext4_ext_calc_credits_for_insert(struct inode *inode,
 
 static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
                                struct ext4_extent *ex,
-                               unsigned long from, unsigned long to)
+                               ext4_lblk_t from, ext4_lblk_t to)
 {
        struct buffer_head *bh;
        unsigned short ee_len =  ext4_ext_get_actual_len(ex);
-       int i;
+       int i, metadata = 0;
 
+       if (S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))
+               metadata = 1;
 #ifdef EXTENTS_STATS
        {
                struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb);
@@ -1738,42 +1787,45 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode,
        if (from >= le32_to_cpu(ex->ee_block)
            && to == le32_to_cpu(ex->ee_block) + ee_len - 1) {
                /* tail removal */
-               unsigned long num;
+               ext4_lblk_t num;
                ext4_fsblk_t start;
+
                num = le32_to_cpu(ex->ee_block) + ee_len - from;
                start = ext_pblock(ex) + ee_len - num;
-               ext_debug("free last %lu blocks starting %llu\n", num, start);
+               ext_debug("free last %u blocks starting %llu\n", num, start);
                for (i = 0; i < num; i++) {
                        bh = sb_find_get_block(inode->i_sb, start + i);
                        ext4_forget(handle, 0, inode, bh, start + i);
                }
-               ext4_free_blocks(handle, inode, start, num);
+               ext4_free_blocks(handle, inode, start, num, metadata);
        } else if (from == le32_to_cpu(ex->ee_block)
                   && to <= le32_to_cpu(ex->ee_block) + ee_len - 1) {
-               printk("strange request: removal %lu-%lu from %u:%u\n",
+               printk(KERN_INFO "strange request: removal %u-%u from %u:%u\n",
                        from, to, le32_to_cpu(ex->ee_block), ee_len);
        } else {
-               printk("strange request: removal(2) %lu-%lu from %u:%u\n",
-                       from, to, le32_to_cpu(ex->ee_block), ee_len);
+               printk(KERN_INFO "strange request: removal(2) "
+                               "%u-%u from %u:%u\n",
+                               from, to, le32_to_cpu(ex->ee_block), ee_len);
        }
        return 0;
 }
 
 static int
 ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
-               struct ext4_ext_path *path, unsigned long start)
+               struct ext4_ext_path *path, ext4_lblk_t start)
 {
        int err = 0, correct_index = 0;
        int depth = ext_depth(inode), credits;
        struct ext4_extent_header *eh;
-       unsigned a, b, block, num;
-       unsigned long ex_ee_block;
+       ext4_lblk_t a, b, block;
+       unsigned num;
+       ext4_lblk_t ex_ee_block;
        unsigned short ex_ee_len;
        unsigned uninitialized = 0;
        struct ext4_extent *ex;
 
        /* the header must be checked already in ext4_ext_remove_space() */
-       ext_debug("truncate since %lu in leaf\n", start);
+       ext_debug("truncate since %u in leaf\n", start);
        if (!path[depth].p_hdr)
                path[depth].p_hdr = ext_block_hdr(path[depth].p_bh);
        eh = path[depth].p_hdr;
@@ -1848,7 +1900,7 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode,
                if (num == 0) {
                        /* this extent is removed; mark slot entirely unused */
                        ext4_ext_store_pblock(ex, 0);
-                       eh->eh_entries = cpu_to_le16(le16_to_cpu(eh->eh_entries)-1);
+                       le16_add_cpu(&eh->eh_entries, -1);
                }
 
                ex->ee_block = cpu_to_le32(block);
@@ -1904,7 +1956,7 @@ ext4_ext_more_to_rm(struct ext4_ext_path *path)
        return 1;
 }
 
-int ext4_ext_remove_space(struct inode *inode, unsigned long start)
+static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start)
 {
        struct super_block *sb = inode->i_sb;
        int depth = ext_depth(inode);
@@ -1912,7 +1964,7 @@ int ext4_ext_remove_space(struct inode *inode, unsigned long start)
        handle_t *handle;
        int i = 0, err = 0;
 
-       ext_debug("truncate since %lu\n", start);
+       ext_debug("truncate since %u\n", start);
 
        /* probably first extent we're gonna free will be last in block */
        handle = ext4_journal_start(inode, depth + 1);
@@ -1925,7 +1977,7 @@ int ext4_ext_remove_space(struct inode *inode, unsigned long start)
         * We start scanning from right side, freeing all the blocks
         * after i_size and walking into the tree depth-wise.
         */
-       path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_KERNEL);
+       path = kzalloc(sizeof(struct ext4_ext_path) * (depth + 1), GFP_NOFS);
        if (path == NULL) {
                ext4_journal_stop(handle);
                return -ENOMEM;
@@ -2084,6 +2136,82 @@ void ext4_ext_release(struct super_block *sb)
 #endif
 }
 
+static void bi_complete(struct bio *bio, int error)
+{
+       complete((struct completion *)bio->bi_private);
+}
+
+/* FIXME!! we need to try to merge to left or right after zero-out  */
+static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
+{
+       int ret = -EIO;
+       struct bio *bio;
+       int blkbits, blocksize;
+       sector_t ee_pblock;
+       struct completion event;
+       unsigned int ee_len, len, done, offset;
+
+
+       blkbits   = inode->i_blkbits;
+       blocksize = inode->i_sb->s_blocksize;
+       ee_len    = ext4_ext_get_actual_len(ex);
+       ee_pblock = ext_pblock(ex);
+
+       /* convert ee_pblock to 512 byte sectors */
+       ee_pblock = ee_pblock << (blkbits - 9);
+
+       while (ee_len > 0) {
+
+               if (ee_len > BIO_MAX_PAGES)
+                       len = BIO_MAX_PAGES;
+               else
+                       len = ee_len;
+
+               bio = bio_alloc(GFP_NOIO, len);
+               if (!bio)
+                       return -ENOMEM;
+               bio->bi_sector = ee_pblock;
+               bio->bi_bdev   = inode->i_sb->s_bdev;
+
+               done = 0;
+               offset = 0;
+               while (done < len) {
+                       ret = bio_add_page(bio, ZERO_PAGE(0),
+                                                       blocksize, offset);
+                       if (ret != blocksize) {
+                               /*
+                                * We can't add any more pages because of
+                                * hardware limitations.  Start a new bio.
+                                */
+                               break;
+                       }
+                       done++;
+                       offset += blocksize;
+                       if (offset >= PAGE_CACHE_SIZE)
+                               offset = 0;
+               }
+
+               init_completion(&event);
+               bio->bi_private = &event;
+               bio->bi_end_io = bi_complete;
+               submit_bio(WRITE, bio);
+               wait_for_completion(&event);
+
+               if (test_bit(BIO_UPTODATE, &bio->bi_flags))
+                       ret = 0;
+               else {
+                       ret = -EIO;
+                       break;
+               }
+               bio_put(bio);
+               ee_len    -= done;
+               ee_pblock += done  << (blkbits - 9);
+       }
+       return ret;
+}
+
+#define EXT4_EXT_ZERO_LEN 7
+
 /*
  * This function is called by ext4_ext_get_blocks() if someone tries to write
  * to an uninitialized extent. It may result in splitting the uninitialized
@@ -2094,17 +2222,19 @@ void ext4_ext_release(struct super_block *sb)
  *   b> Splits in two extents: Write is happening at either end of the extent
  *   c> Splits in three extents: Somone is writing in middle of the extent
  */
-int ext4_ext_convert_to_initialized(handle_t *handle, struct inode *inode,
-                                       struct ext4_ext_path *path,
-                                       ext4_fsblk_t iblock,
-                                       unsigned long max_blocks)
+static int ext4_ext_convert_to_initialized(handle_t *handle,
+                                               struct inode *inode,
+                                               struct ext4_ext_path *path,
+                                               ext4_lblk_t iblock,
+                                               unsigned long max_blocks)
 {
-       struct ext4_extent *ex, newex;
+       struct ext4_extent *ex, newex, orig_ex;
        struct ext4_extent *ex1 = NULL;
        struct ext4_extent *ex2 = NULL;
        struct ext4_extent *ex3 = NULL;
        struct ext4_extent_header *eh;
-       unsigned int allocated, ee_block, ee_len, depth;
+       ext4_lblk_t ee_block;
+       unsigned int allocated, ee_len, depth;
        ext4_fsblk_t newblock;
        int err = 0;
        int ret = 0;
@@ -2117,6 +2247,26 @@ int ext4_ext_convert_to_initialized(handle_t *handle, struct inode *inode,
        allocated = ee_len - (iblock - ee_block);
        newblock = iblock - ee_block + ext_pblock(ex);
        ex2 = ex;
+       orig_ex.ee_block = ex->ee_block;
+       orig_ex.ee_len   = cpu_to_le16(ee_len);
+       ext4_ext_store_pblock(&orig_ex, ext_pblock(ex));
+
+       err = ext4_ext_get_access(handle, inode, path + depth);
+       if (err)
+               goto out;
+       /* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */
+       if (ee_len <= 2*EXT4_EXT_ZERO_LEN) {
+               err =  ext4_ext_zeroout(inode, &orig_ex);
+               if (err)
+                       goto fix_extent_len;
+               /* update the extent length and mark as initialized */
+               ex->ee_block = orig_ex.ee_block;
+               ex->ee_len   = orig_ex.ee_len;
+               ext4_ext_store_pblock(ex, ext_pblock(&orig_ex));
+               ext4_ext_dirty(handle, inode, path + depth);
+               /* zeroed the full extent */
+               return allocated;
+       }
 
        /* ex1: ee_block to iblock - 1 : uninitialized */
        if (iblock > ee_block) {
@@ -2135,33 +2285,139 @@ int ext4_ext_convert_to_initialized(handle_t *handle, struct inode *inode,
        /* ex3: to ee_block + ee_len : uninitialised */
        if (allocated > max_blocks) {
                unsigned int newdepth;
+               /* If extent has less than EXT4_EXT_ZERO_LEN zerout directly */
+               if (allocated <= EXT4_EXT_ZERO_LEN) {
+                       /* Mark first half uninitialized.
+                        * Mark second half initialized and zero out the
+                        * initialized extent
+                        */
+                       ex->ee_block = orig_ex.ee_block;
+                       ex->ee_len   = cpu_to_le16(ee_len - allocated);
+                       ext4_ext_mark_uninitialized(ex);
+                       ext4_ext_store_pblock(ex, ext_pblock(&orig_ex));
+                       ext4_ext_dirty(handle, inode, path + depth);
+
+                       ex3 = &newex;
+                       ex3->ee_block = cpu_to_le32(iblock);
+                       ext4_ext_store_pblock(ex3, newblock);
+                       ex3->ee_len = cpu_to_le16(allocated);
+                       err = ext4_ext_insert_extent(handle, inode, path, ex3);
+                       if (err == -ENOSPC) {
+                               err =  ext4_ext_zeroout(inode, &orig_ex);
+                               if (err)
+                                       goto fix_extent_len;
+                               ex->ee_block = orig_ex.ee_block;
+                               ex->ee_len   = orig_ex.ee_len;
+                               ext4_ext_store_pblock(ex, ext_pblock(&orig_ex));
+                               ext4_ext_dirty(handle, inode, path + depth);
+                               /* zeroed the full extent */
+                               return allocated;
+
+                       } else if (err)
+                               goto fix_extent_len;
+
+                       /*
+                        * We need to zero out the second half because
+                        * an fallocate request can update file size and
+                        * converting the second half to initialized extent
+                        * implies that we can leak some junk data to user
+                        * space.
+                        */
+                       err =  ext4_ext_zeroout(inode, ex3);
+                       if (err) {
+                               /*
+                                * We should actually mark the
+                                * second half as uninit and return error
+                                * Insert would have changed the extent
+                                */
+                               depth = ext_depth(inode);
+                               ext4_ext_drop_refs(path);
+                               path = ext4_ext_find_extent(inode,
+                                                               iblock, path);
+                               if (IS_ERR(path)) {
+                                       err = PTR_ERR(path);
+                                       return err;
+                               }
+                               ex = path[depth].p_ext;
+                               err = ext4_ext_get_access(handle, inode,
+                                                               path + depth);
+                               if (err)
+                                       return err;
+                               ext4_ext_mark_uninitialized(ex);
+                               ext4_ext_dirty(handle, inode, path + depth);
+                               return err;
+                       }
+
+                       /* zeroed the second half */
+                       return allocated;
+               }
                ex3 = &newex;
                ex3->ee_block = cpu_to_le32(iblock + max_blocks);
                ext4_ext_store_pblock(ex3, newblock + max_blocks);
                ex3->ee_len = cpu_to_le16(allocated - max_blocks);
                ext4_ext_mark_uninitialized(ex3);
                err = ext4_ext_insert_extent(handle, inode, path, ex3);
-               if (err)
-                       goto out;
+               if (err == -ENOSPC) {
+                       err =  ext4_ext_zeroout(inode, &orig_ex);
+                       if (err)
+                               goto fix_extent_len;
+                       /* update the extent length and mark as initialized */
+                       ex->ee_block = orig_ex.ee_block;
+                       ex->ee_len   = orig_ex.ee_len;
+                       ext4_ext_store_pblock(ex, ext_pblock(&orig_ex));
+                       ext4_ext_dirty(handle, inode, path + depth);
+                       /* zeroed the full extent */
+                       return allocated;
+
+               } else if (err)
+                       goto fix_extent_len;
                /*
                 * The depth, and hence eh & ex might change
                 * as part of the insert above.
                 */
                newdepth = ext_depth(inode);
+               /*
+                * update the extent length after successfull insert of the
+                * split extent
+                */
+               orig_ex.ee_len = cpu_to_le16(ee_len -
+                                               ext4_ext_get_actual_len(ex3));
                if (newdepth != depth) {
                        depth = newdepth;
-                       path = ext4_ext_find_extent(inode, iblock, NULL);
+                       ext4_ext_drop_refs(path);
+                       path = ext4_ext_find_extent(inode, iblock, path);
                        if (IS_ERR(path)) {
                                err = PTR_ERR(path);
-                               path = NULL;
                                goto out;
                        }
                        eh = path[depth].p_hdr;
                        ex = path[depth].p_ext;
                        if (ex2 != &newex)
                                ex2 = ex;
+
+                       err = ext4_ext_get_access(handle, inode, path + depth);
+                       if (err)
+                               goto out;
                }
                allocated = max_blocks;
+
+               /* If extent has less than EXT4_EXT_ZERO_LEN and we are trying
+                * to insert a extent in the middle zerout directly
+                * otherwise give the extent a chance to merge to left
+                */
+               if (le16_to_cpu(orig_ex.ee_len) <= EXT4_EXT_ZERO_LEN &&
+                                                       iblock != ee_block) {
+                       err =  ext4_ext_zeroout(inode, &orig_ex);
+                       if (err)
+                               goto fix_extent_len;
+                       /* update the extent length and mark as initialized */
+                       ex->ee_block = orig_ex.ee_block;
+                       ex->ee_len   = orig_ex.ee_len;
+                       ext4_ext_store_pblock(ex, ext_pblock(&orig_ex));
+                       ext4_ext_dirty(handle, inode, path + depth);
+                       /* zero out the first half */
+                       return allocated;
+               }
        }
        /*
         * If there was a change of depth as part of the
@@ -2180,9 +2436,6 @@ int ext4_ext_convert_to_initialized(handle_t *handle, struct inode *inode,
        ex2->ee_len = cpu_to_le16(allocated);
        if (ex2 != ex)
                goto insert;
-       err = ext4_ext_get_access(handle, inode, path + depth);
-       if (err)
-               goto out;
        /*
         * New (initialized) extent starts from the first block
         * in the current extent. i.e., ex2 == ex
@@ -2221,12 +2474,51 @@ int ext4_ext_convert_to_initialized(handle_t *handle, struct inode *inode,
        goto out;
 insert:
        err = ext4_ext_insert_extent(handle, inode, path, &newex);
+       if (err == -ENOSPC) {
+               err =  ext4_ext_zeroout(inode, &orig_ex);
+               if (err)
+                       goto fix_extent_len;
+               /* update the extent length and mark as initialized */
+               ex->ee_block = orig_ex.ee_block;
+               ex->ee_len   = orig_ex.ee_len;
+               ext4_ext_store_pblock(ex, ext_pblock(&orig_ex));
+               ext4_ext_dirty(handle, inode, path + depth);
+               /* zero out the first half */
+               return allocated;
+       } else if (err)
+               goto fix_extent_len;
 out:
        return err ? err : allocated;
+
+fix_extent_len:
+       ex->ee_block = orig_ex.ee_block;
+       ex->ee_len   = orig_ex.ee_len;
+       ext4_ext_store_pblock(ex, ext_pblock(&orig_ex));
+       ext4_ext_mark_uninitialized(ex);
+       ext4_ext_dirty(handle, inode, path + depth);
+       return err;
 }
 
+/*
+ * Block allocation/map/preallocation routine for extents based files
+ *
+ *
+ * Need to be called with
+ * down_read(&EXT4_I(inode)->i_data_sem) if not allocating file system block
+ * (ie, create is zero). Otherwise down_write(&EXT4_I(inode)->i_data_sem)
+ *
+ * return > 0, number of of blocks already mapped/allocated
+ *          if create == 0 and these are pre-allocated blocks
+ *             buffer head is unmapped
+ *          otherwise blocks are mapped
+ *
+ * return = 0, if plain look up failed (blocks have not been allocated)
+ *          buffer head is unmapped
+ *
+ * return < 0, error case.
+ */
 int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
-                       ext4_fsblk_t iblock,
+                       ext4_lblk_t iblock,
                        unsigned long max_blocks, struct buffer_head *bh_result,
                        int create, int extend_disksize)
 {
@@ -2236,11 +2528,11 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
        ext4_fsblk_t goal, newblock;
        int err = 0, depth, ret;
        unsigned long allocated = 0;
+       struct ext4_allocation_request ar;
 
        __clear_bit(BH_New, &bh_result->b_state);
-       ext_debug("blocks %d/%lu requested for inode %u\n", (int) iblock,
-                       max_blocks, (unsigned) inode->i_ino);
-       mutex_lock(&EXT4_I(inode)->truncate_mutex);
+       ext_debug("blocks %u/%lu requested for inode %u\n",
+                       iblock, max_blocks, inode->i_ino);
 
        /* check in cache */
        goal = ext4_ext_in_cache(inode, iblock, &newex);
@@ -2260,7 +2552,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
                                   - le32_to_cpu(newex.ee_block)
                                   + ext_pblock(&newex);
                        /* number of remaining blocks in the extent */
-                       allocated = le16_to_cpu(newex.ee_len) -
+                       allocated = ext4_ext_get_actual_len(&newex) -
                                        (iblock - le32_to_cpu(newex.ee_block));
                        goto out;
                } else {
@@ -2288,7 +2580,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 
        ex = path[depth].p_ext;
        if (ex) {
-               unsigned long ee_block = le32_to_cpu(ex->ee_block);
+               ext4_lblk_t ee_block = le32_to_cpu(ex->ee_block);
                ext4_fsblk_t ee_start = ext_pblock(ex);
                unsigned short ee_len;
 
@@ -2302,7 +2594,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
                        newblock = iblock - ee_block + ee_start;
                        /* number of remaining blocks in the extent */
                        allocated = ee_len - (iblock - ee_block);
-                       ext_debug("%d fit into %lu:%d -> %llu\n", (int) iblock,
+                       ext_debug("%u fit into %lu:%d -> %llu\n", iblock,
                                        ee_block, ee_len, newblock);
 
                        /* Do not put uninitialized extent in the cache */
@@ -2314,15 +2606,28 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
                        }
                        if (create == EXT4_CREATE_UNINITIALIZED_EXT)
                                goto out;
-                       if (!create)
+                       if (!create) {
+                               /*
+                                * We have blocks reserved already.  We
+                                * return allocated blocks so that delalloc
+                                * won't do block reservation for us.  But
+                                * the buffer head will be unmapped so that
+                                * a read from the block returns 0s.
+                                */
+                               if (allocated > max_blocks)
+                                       allocated = max_blocks;
+                               /* mark the buffer unwritten */
+                               __set_bit(BH_Unwritten, &bh_result->b_state);
                                goto out2;
+                       }
 
                        ret = ext4_ext_convert_to_initialized(handle, inode,
                                                                path, iblock,
                                                                max_blocks);
-                       if (ret <= 0)
+                       if (ret <= 0) {
+                               err = ret;
                                goto out2;
-                       else
+                       else
                                allocated = ret;
                        goto outnew;
                }
@@ -2347,8 +2652,15 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
        if (S_ISREG(inode->i_mode) && (!EXT4_I(inode)->i_block_alloc_info))
                ext4_init_block_alloc_info(inode);
 
-       /* allocate new block */
-       goal = ext4_ext_find_goal(inode, path, iblock);
+       /* find neighbour allocated blocks */
+       ar.lleft = iblock;
+       err = ext4_ext_search_left(inode, path, &ar.lleft, &ar.pleft);
+       if (err)
+               goto out2;
+       ar.lright = iblock;
+       err = ext4_ext_search_right(inode, path, &ar.lright, &ar.pright);
+       if (err)
+               goto out2;
 
        /*
         * See if request is beyond maximum number of blocks we can have in
@@ -2368,10 +2680,21 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
        newex.ee_len = cpu_to_le16(max_blocks);
        err = ext4_ext_check_overlap(inode, &newex, path);
        if (err)
-               allocated = le16_to_cpu(newex.ee_len);
+               allocated = ext4_ext_get_actual_len(&newex);
        else
                allocated = max_blocks;
-       newblock = ext4_new_blocks(handle, inode, goal, &allocated, &err);
+
+       /* allocate new block */
+       ar.inode = inode;
+       ar.goal = ext4_ext_find_goal(inode, path, iblock);
+       ar.logical = iblock;
+       ar.len = allocated;
+       if (S_ISREG(inode->i_mode))
+               ar.flags = EXT4_MB_HINT_DATA;
+       else
+               /* disable in-core preallocation for non-regular files */
+               ar.flags = 0;
+       newblock = ext4_mb_new_blocks(handle, &ar, &err);
        if (!newblock)
                goto out2;
        ext_debug("allocate new block: goal %llu, found %llu/%lu\n",
@@ -2379,14 +2702,17 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 
        /* try to insert new extent into found leaf and return */
        ext4_ext_store_pblock(&newex, newblock);
-       newex.ee_len = cpu_to_le16(allocated);
+       newex.ee_len = cpu_to_le16(ar.len);
        if (create == EXT4_CREATE_UNINITIALIZED_EXT)  /* Mark uninitialized */
                ext4_ext_mark_uninitialized(&newex);
        err = ext4_ext_insert_extent(handle, inode, path, &newex);
        if (err) {
                /* free data blocks we just allocated */
+               /* not a good idea to call discard here directly,
+                * but otherwise we'd need to call it every free() */
+               ext4_mb_discard_inode_preallocations(inode);
                ext4_free_blocks(handle, inode, ext_pblock(&newex),
-                                       le16_to_cpu(newex.ee_len));
+                                       ext4_ext_get_actual_len(&newex), 0);
                goto out2;
        }
 
@@ -2395,6 +2721,7 @@ int ext4_ext_get_blocks(handle_t *handle, struct inode *inode,
 
        /* previous routine could use block we allocated */
        newblock = ext_pblock(&newex);
+       allocated = ext4_ext_get_actual_len(&newex);
 outnew:
        __set_bit(BH_New, &bh_result->b_state);
 
@@ -2414,8 +2741,6 @@ out2:
                ext4_ext_drop_refs(path);
                kfree(path);
        }
-       mutex_unlock(&EXT4_I(inode)->truncate_mutex);
-
        return err ? err : allocated;
 }
 
@@ -2423,7 +2748,7 @@ void ext4_ext_truncate(struct inode * inode, struct page *page)
 {
        struct address_space *mapping = inode->i_mapping;
        struct super_block *sb = inode->i_sb;
-       unsigned long last_block;
+       ext4_lblk_t last_block;
        handle_t *handle;
        int err = 0;
 
@@ -2445,9 +2770,11 @@ void ext4_ext_truncate(struct inode * inode, struct page *page)
        if (page)
                ext4_block_truncate_page(handle, page, mapping, inode->i_size);
 
-       mutex_lock(&EXT4_I(inode)->truncate_mutex);
+       down_write(&EXT4_I(inode)->i_data_sem);
        ext4_ext_invalidate_cache(inode);
 
+       ext4_mb_discard_inode_preallocations(inode);
+
        /*
         * TODO: optimization is possible here.
         * Probably we need not scan at all,
@@ -2481,7 +2808,7 @@ out_stop:
        if (inode->i_nlink)
                ext4_orphan_del(handle, inode);
 
-       mutex_unlock(&EXT4_I(inode)->truncate_mutex);
+       up_write(&EXT4_I(inode)->i_data_sem);
        ext4_journal_stop(handle);
 }
 
@@ -2506,6 +2833,28 @@ int ext4_ext_writepage_trans_blocks(struct inode *inode, int num)
        return needed;
 }
 
+static void ext4_falloc_update_inode(struct inode *inode,
+                               int mode, loff_t new_size, int update_ctime)
+{
+       struct timespec now;
+
+       if (update_ctime) {
+               now = current_fs_time(inode->i_sb);
+               if (!timespec_equal(&inode->i_ctime, &now))
+                       inode->i_ctime = now;
+       }
+       /*
+        * Update only when preallocation was requested beyond
+        * the file size.
+        */
+       if (!(mode & FALLOC_FL_KEEP_SIZE) &&
+                               new_size > i_size_read(inode)) {
+               i_size_write(inode, new_size);
+               EXT4_I(inode)->i_disksize = new_size;
+       }
+
+}
+
 /*
  * preallocate space for a file. This implements ext4's fallocate inode
  * operation, which gets called from sys_fallocate system call.
@@ -2516,8 +2865,9 @@ int ext4_ext_writepage_trans_blocks(struct inode *inode, int num)
 long ext4_fallocate(struct inode *inode, int mode, loff_t offset, loff_t len)
 {
        handle_t *handle;
-       ext4_fsblk_t block, max_blocks;
-       ext4_fsblk_t nblocks = 0;
+       ext4_lblk_t block;
+       loff_t new_size;
+       unsigned long max_blocks;
        int ret = 0;
        int ret2 = 0;
        int retries = 0;
@@ -2536,14 +2886,18 @@ long ext4_fallocate(struct inode *inode, int mode, loff_t offset, loff_t len)
                return -ENODEV;
 
        block = offset >> blkbits;
+       /*
+        * We can't just convert len to max_blocks because
+        * If blocksize = 4096 offset = 3072 and len = 2048
+        */
        max_blocks = (EXT4_BLOCK_ALIGN(len + offset, blkbits) >> blkbits)
-                       - block;
-
+                                                       - block;
        /*
         * credits to insert 1 extent into extent tree + buffers to be able to
         * modify 1 super block, 1 block bitmap and 1 group descriptor.
         */
        credits = EXT4_DATA_TRANS_BLOCKS(inode->i_sb) + 3;
+       mutex_lock(&inode->i_mutex);
 retry:
        while (ret >= 0 && ret < max_blocks) {
                block = block + ret;
@@ -2553,79 +2907,39 @@ retry:
                        ret = PTR_ERR(handle);
                        break;
                }
-
-               ret = ext4_ext_get_blocks(handle, inode, block,
+               ret = ext4_get_blocks_wrap(handle, inode, block,
                                          max_blocks, &map_bh,
                                          EXT4_CREATE_UNINITIALIZED_EXT, 0);
-               WARN_ON(!ret);
-               if (!ret) {
-                       ext4_error(inode->i_sb, "ext4_fallocate",
-                                  "ext4_ext_get_blocks returned 0! inode#%lu"
-                                  ", block=%llu, max_blocks=%llu",
-                                  inode->i_ino, block, max_blocks);
-                       ret = -EIO;
+               if (ret <= 0) {
+#ifdef EXT4FS_DEBUG
+                       WARN_ON(ret <= 0);
+                       printk(KERN_ERR "%s: ext4_ext_get_blocks "
+                                   "returned error inode#%lu, block=%u, "
+                                   "max_blocks=%lu", __func__,
+                                   inode->i_ino, block, max_blocks);
+#endif
                        ext4_mark_inode_dirty(handle, inode);
                        ret2 = ext4_journal_stop(handle);
                        break;
                }
-               if (ret > 0) {
-                       /* check wrap through sign-bit/zero here */
-                       if ((block + ret) < 0 || (block + ret) < block) {
-                               ret = -EIO;
-                               ext4_mark_inode_dirty(handle, inode);
-                               ret2 = ext4_journal_stop(handle);
-                               break;
-                       }
-                       if (buffer_new(&map_bh) && ((block + ret) >
-                           (EXT4_BLOCK_ALIGN(i_size_read(inode), blkbits)
-                           >> blkbits)))
-                                       nblocks = nblocks + ret;
-               }
-
-               /* Update ctime if new blocks get allocated */
-               if (nblocks) {
-                       struct timespec now;
-
-                       now = current_fs_time(inode->i_sb);
-                       if (!timespec_equal(&inode->i_ctime, &now))
-                               inode->i_ctime = now;
-               }
+               if ((block + ret) >= (EXT4_BLOCK_ALIGN(offset + len,
+                                               blkbits) >> blkbits))
+                       new_size = offset + len;
+               else
+                       new_size = (block + ret) << blkbits;
 
+               ext4_falloc_update_inode(inode, mode, new_size,
+                                               buffer_new(&map_bh));
                ext4_mark_inode_dirty(handle, inode);
                ret2 = ext4_journal_stop(handle);
                if (ret2)
                        break;
        }
-
-       if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries))
+       if (ret == -ENOSPC &&
+                       ext4_should_retry_alloc(inode->i_sb, &retries)) {
+               ret = 0;
                goto retry;
-
-       /*
-        * Time to update the file size.
-        * Update only when preallocation was requested beyond the file size.
-        */
-       if (!(mode & FALLOC_FL_KEEP_SIZE) &&
-           (offset + len) > i_size_read(inode)) {
-               if (ret > 0) {
-                       /*
-                        * if no error, we assume preallocation succeeded
-                        * completely
-                        */
-                       mutex_lock(&inode->i_mutex);
-                       i_size_write(inode, offset + len);
-                       EXT4_I(inode)->i_disksize = i_size_read(inode);
-                       mutex_unlock(&inode->i_mutex);
-               } else if (ret < 0 && nblocks) {
-                       /* Handle partial allocation scenario */
-                       loff_t newsize;
-
-                       mutex_lock(&inode->i_mutex);
-                       newsize  = (nblocks << blkbits) + i_size_read(inode);
-                       i_size_write(inode, EXT4_BLOCK_ALIGN(newsize, blkbits));
-                       EXT4_I(inode)->i_disksize = i_size_read(inode);
-                       mutex_unlock(&inode->i_mutex);
-               }
        }
-
+       mutex_unlock(&inode->i_mutex);
        return ret > 0 ? ret2 : ret;
 }