ocfs2: support for removing file regions
authorMark Fasheh <mark.fasheh@oracle.com>
Tue, 3 Jul 2007 20:34:11 +0000 (13:34 -0700)
committerMark Fasheh <mark.fasheh@oracle.com>
Wed, 11 Jul 2007 00:32:08 +0000 (17:32 -0700)
Provide an internal interface for the removal of arbitrary file regions.

ocfs2_remove_inode_range() takes a byte range within a file and will remove
existing extents within that range. Partial clusters will be zeroed so that
any read from within the region will return zeros.

Signed-off-by: Mark Fasheh <mark.fasheh@oracle.com>
fs/ocfs2/alloc.c
fs/ocfs2/alloc.h
fs/ocfs2/file.c
fs/ocfs2/journal.h

index df186d2e82485b407697de04b4982b2507a03c08..f5e11f4fa952a424a5259d766a03c1d449c8ec31 100644 (file)
@@ -4373,10 +4373,10 @@ out:
        return ret;
 }
 
-static int ocfs2_remove_extent(struct inode *inode, struct buffer_head *di_bh,
-                              u32 cpos, u32 len, handle_t *handle,
-                              struct ocfs2_alloc_context *meta_ac,
-                              struct ocfs2_cached_dealloc_ctxt *dealloc)
+int ocfs2_remove_extent(struct inode *inode, struct buffer_head *di_bh,
+                       u32 cpos, u32 len, handle_t *handle,
+                       struct ocfs2_alloc_context *meta_ac,
+                       struct ocfs2_cached_dealloc_ctxt *dealloc)
 {
        int ret, index;
        u32 rec_range, trunc_range;
@@ -4506,7 +4506,7 @@ out:
        return ret;
 }
 
-static inline int ocfs2_truncate_log_needs_flush(struct ocfs2_super *osb)
+int ocfs2_truncate_log_needs_flush(struct ocfs2_super *osb)
 {
        struct buffer_head *tl_bh = osb->osb_tl_bh;
        struct ocfs2_dinode *di;
@@ -4539,10 +4539,10 @@ static int ocfs2_truncate_log_can_coalesce(struct ocfs2_truncate_log *tl,
        return current_tail == new_start;
 }
 
-static int ocfs2_truncate_log_append(struct ocfs2_super *osb,
-                                    handle_t *handle,
-                                    u64 start_blk,
-                                    unsigned int num_clusters)
+int ocfs2_truncate_log_append(struct ocfs2_super *osb,
+                             handle_t *handle,
+                             u64 start_blk,
+                             unsigned int num_clusters)
 {
        int status, index;
        unsigned int start_cluster, tl_count;
@@ -4698,7 +4698,7 @@ bail:
 }
 
 /* Expects you to already be holding tl_inode->i_mutex */
-static int __ocfs2_flush_truncate_log(struct ocfs2_super *osb)
+int __ocfs2_flush_truncate_log(struct ocfs2_super *osb)
 {
        int status;
        unsigned int num_to_flush;
index 752ef860873d6c64ef409c2d1f008dcd9f5693c7..990df48ae8d361459ee27b42cccfb7b0f935c222 100644 (file)
@@ -41,6 +41,10 @@ int ocfs2_mark_extent_written(struct inode *inode, struct buffer_head *di_bh,
                              handle_t *handle, u32 cpos, u32 len, u32 phys,
                              struct ocfs2_alloc_context *meta_ac,
                              struct ocfs2_cached_dealloc_ctxt *dealloc);
+int ocfs2_remove_extent(struct inode *inode, struct buffer_head *di_bh,
+                       u32 cpos, u32 len, handle_t *handle,
+                       struct ocfs2_alloc_context *meta_ac,
+                       struct ocfs2_cached_dealloc_ctxt *dealloc);
 int ocfs2_num_free_extents(struct ocfs2_super *osb,
                           struct inode *inode,
                           struct ocfs2_dinode *fe);
@@ -68,6 +72,12 @@ int ocfs2_begin_truncate_log_recovery(struct ocfs2_super *osb,
                                      struct ocfs2_dinode **tl_copy);
 int ocfs2_complete_truncate_log_recovery(struct ocfs2_super *osb,
                                         struct ocfs2_dinode *tl_copy);
+int ocfs2_truncate_log_needs_flush(struct ocfs2_super *osb);
+int ocfs2_truncate_log_append(struct ocfs2_super *osb,
+                             handle_t *handle,
+                             u64 start_blk,
+                             unsigned int num_clusters);
+int __ocfs2_flush_truncate_log(struct ocfs2_super *osb);
 
 /*
  * Process local structure which describes the block unlinks done
index f0a6b1330a6ecdade605632118451e72bb90edf3..11f7cf9f251122a84803a77d8f2334e201a10bee 100644 (file)
@@ -541,12 +541,15 @@ int ocfs2_lock_allocators(struct inode *inode, struct ocfs2_dinode *di,
                          struct ocfs2_alloc_context **data_ac,
                          struct ocfs2_alloc_context **meta_ac)
 {
-       int ret, num_free_extents;
+       int ret = 0, num_free_extents;
        unsigned int max_recs_needed = clusters_to_add + 2 * extents_to_split;
        struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
 
        *meta_ac = NULL;
-       *data_ac = NULL;
+       if (data_ac)
+               *data_ac = NULL;
+
+       BUG_ON(clusters_to_add != 0 && data_ac == NULL);
 
        mlog(0, "extend inode %llu, i_size = %lld, di->i_clusters = %u, "
             "clusters_to_add = %u, extents_to_split = %u\n",
@@ -583,6 +586,9 @@ int ocfs2_lock_allocators(struct inode *inode, struct ocfs2_dinode *di,
                }
        }
 
+       if (clusters_to_add == 0)
+               goto out;
+
        ret = ocfs2_reserve_clusters(osb, clusters_to_add, data_ac);
        if (ret < 0) {
                if (ret != -ENOSPC)
@@ -1252,6 +1258,238 @@ out:
        return ret;
 }
 
+static int __ocfs2_remove_inode_range(struct inode *inode,
+                                     struct buffer_head *di_bh,
+                                     u32 cpos, u32 phys_cpos, u32 len,
+                                     struct ocfs2_cached_dealloc_ctxt *dealloc)
+{
+       int ret;
+       u64 phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos);
+       struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+       struct inode *tl_inode = osb->osb_tl_inode;
+       handle_t *handle;
+       struct ocfs2_alloc_context *meta_ac = NULL;
+       struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data;
+
+       ret = ocfs2_lock_allocators(inode, di, 0, 1, NULL, &meta_ac);
+       if (ret) {
+               mlog_errno(ret);
+               return ret;
+       }
+
+       mutex_lock(&tl_inode->i_mutex);
+
+       if (ocfs2_truncate_log_needs_flush(osb)) {
+               ret = __ocfs2_flush_truncate_log(osb);
+               if (ret < 0) {
+                       mlog_errno(ret);
+                       goto out;
+               }
+       }
+
+       handle = ocfs2_start_trans(osb, OCFS2_REMOVE_EXTENT_CREDITS);
+       if (handle == NULL) {
+               ret = -ENOMEM;
+               mlog_errno(ret);
+               goto out;
+       }
+
+       ret = ocfs2_journal_access(handle, inode, di_bh,
+                                  OCFS2_JOURNAL_ACCESS_WRITE);
+       if (ret) {
+               mlog_errno(ret);
+               goto out;
+       }
+
+       ret = ocfs2_remove_extent(inode, di_bh, cpos, len, handle, meta_ac,
+                                 dealloc);
+       if (ret) {
+               mlog_errno(ret);
+               goto out_commit;
+       }
+
+       OCFS2_I(inode)->ip_clusters -= len;
+       di->i_clusters = cpu_to_le32(OCFS2_I(inode)->ip_clusters);
+
+       ret = ocfs2_journal_dirty(handle, di_bh);
+       if (ret) {
+               mlog_errno(ret);
+               goto out_commit;
+       }
+
+       ret = ocfs2_truncate_log_append(osb, handle, phys_blkno, len);
+       if (ret)
+               mlog_errno(ret);
+
+out_commit:
+       ocfs2_commit_trans(osb, handle);
+out:
+       mutex_unlock(&tl_inode->i_mutex);
+
+       if (meta_ac)
+               ocfs2_free_alloc_context(meta_ac);
+
+       return ret;
+}
+
+/*
+ * Truncate a byte range, avoiding pages within partial clusters. This
+ * preserves those pages for the zeroing code to write to.
+ */
+static void ocfs2_truncate_cluster_pages(struct inode *inode, u64 byte_start,
+                                        u64 byte_len)
+{
+       struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+       loff_t start, end;
+       struct address_space *mapping = inode->i_mapping;
+
+       start = (loff_t)ocfs2_align_bytes_to_clusters(inode->i_sb, byte_start);
+       end = byte_start + byte_len;
+       end = end & ~(osb->s_clustersize - 1);
+
+       if (start < end) {
+               unmap_mapping_range(mapping, start, end - start, 0);
+               truncate_inode_pages_range(mapping, start, end - 1);
+       }
+}
+
+static int ocfs2_zero_partial_clusters(struct inode *inode,
+                                      u64 start, u64 len)
+{
+       int ret = 0;
+       u64 tmpend, end = start + len;
+       struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+       unsigned int csize = osb->s_clustersize;
+       handle_t *handle;
+
+       /*
+        * The "start" and "end" values are NOT necessarily part of
+        * the range whose allocation is being deleted. Rather, this
+        * is what the user passed in with the request. We must zero
+        * partial clusters here. There's no need to worry about
+        * physical allocation - the zeroing code knows to skip holes.
+        */
+       mlog(0, "byte start: %llu, end: %llu\n",
+            (unsigned long long)start, (unsigned long long)end);
+
+       /*
+        * If both edges are on a cluster boundary then there's no
+        * zeroing required as the region is part of the allocation to
+        * be truncated.
+        */
+       if ((start & (csize - 1)) == 0 && (end & (csize - 1)) == 0)
+               goto out;
+
+       handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
+       if (handle == NULL) {
+               ret = -ENOMEM;
+               mlog_errno(ret);
+               goto out;
+       }
+
+       /*
+        * We want to get the byte offset of the end of the 1st cluster.
+        */
+       tmpend = (u64)osb->s_clustersize + (start & ~(osb->s_clustersize - 1));
+       if (tmpend > end)
+               tmpend = end;
+
+       mlog(0, "1st range: start: %llu, tmpend: %llu\n",
+            (unsigned long long)start, (unsigned long long)tmpend);
+
+       ret = ocfs2_zero_range_for_truncate(inode, handle, start, tmpend);
+       if (ret)
+               mlog_errno(ret);
+
+       if (tmpend < end) {
+               /*
+                * This may make start and end equal, but the zeroing
+                * code will skip any work in that case so there's no
+                * need to catch it up here.
+                */
+               start = end & ~(osb->s_clustersize - 1);
+
+               mlog(0, "2nd range: start: %llu, end: %llu\n",
+                    (unsigned long long)start, (unsigned long long)end);
+
+               ret = ocfs2_zero_range_for_truncate(inode, handle, start, end);
+               if (ret)
+                       mlog_errno(ret);
+       }
+
+       ocfs2_commit_trans(osb, handle);
+out:
+       return ret;
+}
+
+static int ocfs2_remove_inode_range(struct inode *inode,
+                                   struct buffer_head *di_bh, u64 byte_start,
+                                   u64 byte_len)
+{
+       int ret = 0;
+       u32 trunc_start, trunc_len, cpos, phys_cpos, alloc_size;
+       struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+       struct ocfs2_cached_dealloc_ctxt dealloc;
+
+       ocfs2_init_dealloc_ctxt(&dealloc);
+
+       if (byte_len == 0)
+               return 0;
+
+       trunc_start = ocfs2_clusters_for_bytes(osb->sb, byte_start);
+       trunc_len = (byte_start + byte_len) >> osb->s_clustersize_bits;
+       if (trunc_len >= trunc_start)
+               trunc_len -= trunc_start;
+       else
+               trunc_len = 0;
+
+       mlog(0, "Inode: %llu, start: %llu, len: %llu, cstart: %u, clen: %u\n",
+            (unsigned long long)OCFS2_I(inode)->ip_blkno,
+            (unsigned long long)byte_start,
+            (unsigned long long)byte_len, trunc_start, trunc_len);
+
+       ret = ocfs2_zero_partial_clusters(inode, byte_start, byte_len);
+       if (ret) {
+               mlog_errno(ret);
+               goto out;
+       }
+
+       cpos = trunc_start;
+       while (trunc_len) {
+               ret = ocfs2_get_clusters(inode, cpos, &phys_cpos,
+                                        &alloc_size, NULL);
+               if (ret) {
+                       mlog_errno(ret);
+                       goto out;
+               }
+
+               if (alloc_size > trunc_len)
+                       alloc_size = trunc_len;
+
+               /* Only do work for non-holes */
+               if (phys_cpos != 0) {
+                       ret = __ocfs2_remove_inode_range(inode, di_bh, cpos,
+                                                        phys_cpos, alloc_size,
+                                                        &dealloc);
+                       if (ret) {
+                               mlog_errno(ret);
+                               goto out;
+                       }
+               }
+
+               cpos += alloc_size;
+               trunc_len -= alloc_size;
+       }
+
+       ocfs2_truncate_cluster_pages(inode, byte_start, byte_len);
+
+out:
+       ocfs2_schedule_truncate_log_flush(osb, 1);
+       ocfs2_run_deallocs(osb, &dealloc);
+
+       return ret;
+}
+
 static int ocfs2_prepare_inode_for_write(struct dentry *dentry,
                                         loff_t *ppos,
                                         size_t count,
index 3db5de4506daf9510ab8170d39161162a1264941..ce60aab013aa2f7645fddb317b9a8c342654a50b 100644 (file)
@@ -289,6 +289,8 @@ int                  ocfs2_journal_dirty_data(handle_t *handle,
 #define OCFS2_TRUNCATE_LOG_FLUSH_ONE_REC (OCFS2_SUBALLOC_FREE                \
                                         + OCFS2_TRUNCATE_LOG_UPDATE)
 
+#define OCFS2_REMOVE_EXTENT_CREDITS (OCFS2_TRUNCATE_LOG_UPDATE + OCFS2_INODE_UPDATE_CREDITS)
+
 /* data block for new dir/symlink, 2 for bitmap updates (bitmap fe +
  * bitmap block for the new bit) */
 #define OCFS2_DIR_LINK_ADDITIONAL_CREDITS (1 + 2)