* might now be discovering a truncate that hit on another node.
* block_read_full_page->get_block freaks out if it is asked to read
* beyond the end of a file, so we check here. Callers
- * (generic_file_read, fault->nopage) are clever enough to check i_size
+ * (generic_file_read, vm_ops->fault) are clever enough to check i_size
* and notice that the page they just read isn't needed.
*
* XXX sys_readahead() seems to get that wrong?
bh = head;
block_start = 0;
do {
- void *kaddr;
-
block_end = block_start + bsize;
if (block_end <= from)
goto next_bh;
if (block_start >= to)
break;
- kaddr = kmap_atomic(page, KM_USER0);
- memset(kaddr+block_start, 0, bh->b_size);
- flush_dcache_page(page);
- kunmap_atomic(kaddr, KM_USER0);
+ zero_user_page(page, block_start, bh->b_size, KM_USER0);
set_buffer_uptodate(bh);
mark_buffer_dirty(bh);
* filled.
*/
unsigned c_new;
+ unsigned c_unwritten;
};
+static inline int ocfs2_should_zero_cluster(struct ocfs2_write_cluster_desc *d)
+{
+ return d->c_new || d->c_unwritten;
+}
+
struct ocfs2_write_ctxt {
/* Logical cluster position / len of write */
u32 w_cpos;
handle_t *w_handle;
struct buffer_head *w_di_bh;
+
+ struct ocfs2_cached_dealloc_ctxt w_dealloc;
};
static void ocfs2_free_write_ctxt(struct ocfs2_write_ctxt *wc)
struct ocfs2_super *osb, loff_t pos,
unsigned len, struct buffer_head *di_bh)
{
+ u32 cend;
struct ocfs2_write_ctxt *wc;
wc = kzalloc(sizeof(struct ocfs2_write_ctxt), GFP_NOFS);
return -ENOMEM;
wc->w_cpos = pos >> osb->s_clustersize_bits;
- wc->w_clen = ocfs2_clusters_for_bytes(osb->sb, len);
+ cend = (pos + len - 1) >> osb->s_clustersize_bits;
+ wc->w_clen = cend - wc->w_cpos + 1;
get_bh(di_bh);
wc->w_di_bh = di_bh;
else
wc->w_large_pages = 0;
+ ocfs2_init_dealloc_ctxt(&wc->w_dealloc);
+
*wcp = wc;
return 0;
if (block_end > from && block_start < to) {
if (!PageUptodate(page)) {
unsigned start, end;
- void *kaddr;
start = max(from, block_start);
end = min(to, block_end);
- kaddr = kmap_atomic(page, KM_USER0);
- memset(kaddr+start, 0, end - start);
- flush_dcache_page(page);
- kunmap_atomic(kaddr, KM_USER0);
+ zero_user_page(page, start, end - start, KM_USER0);
set_buffer_uptodate(bh);
}
loff_t user_pos, unsigned user_len)
{
int i;
- unsigned from, to;
+ unsigned from = user_pos & (PAGE_CACHE_SIZE - 1),
+ to = user_pos + user_len;
struct page *tmppage;
- ocfs2_zero_new_buffers(wc->w_target_page, user_pos, user_len);
-
- if (wc->w_large_pages) {
- from = wc->w_target_from;
- to = wc->w_target_to;
- } else {
- from = 0;
- to = PAGE_CACHE_SIZE;
- }
+ ocfs2_zero_new_buffers(wc->w_target_page, from, to);
for(i = 0; i < wc->w_num_pages; i++) {
tmppage = wc->w_pages[i];
map_from = cluster_start;
map_to = cluster_end;
}
-
- wc->w_target_from = map_from;
- wc->w_target_to = map_to;
} else {
/*
* If we haven't allocated the new page yet, we
* Prepare a single cluster for write one cluster into the file.
*/
static int ocfs2_write_cluster(struct address_space *mapping,
- u32 phys, struct ocfs2_alloc_context *data_ac,
+ u32 phys, unsigned int unwritten,
+ struct ocfs2_alloc_context *data_ac,
struct ocfs2_alloc_context *meta_ac,
struct ocfs2_write_ctxt *wc, u32 cpos,
loff_t user_pos, unsigned user_len)
{
- int ret, i, new;
+ int ret, i, new, should_zero = 0;
u64 v_blkno, p_blkno;
struct inode *inode = mapping->host;
new = phys == 0 ? 1 : 0;
+ if (new || unwritten)
+ should_zero = 1;
if (new) {
u32 tmp_pos;
*/
tmp_pos = cpos;
ret = ocfs2_do_extend_allocation(OCFS2_SB(inode->i_sb), inode,
- &tmp_pos, 1, wc->w_di_bh,
+ &tmp_pos, 1, 0, wc->w_di_bh,
wc->w_handle, data_ac,
meta_ac, NULL);
/*
mlog_errno(ret);
goto out;
}
+ } else if (unwritten) {
+ ret = ocfs2_mark_extent_written(inode, wc->w_di_bh,
+ wc->w_handle, cpos, 1, phys,
+ meta_ac, &wc->w_dealloc);
+ if (ret < 0) {
+ mlog_errno(ret);
+ goto out;
+ }
+ }
+ if (should_zero)
v_blkno = ocfs2_clusters_to_blocks(inode->i_sb, cpos);
- } else {
+ else
v_blkno = user_pos >> inode->i_sb->s_blocksize_bits;
- }
/*
* The only reason this should fail is due to an inability to
tmpret = ocfs2_prepare_page_for_write(inode, &p_blkno, wc,
wc->w_pages[i], cpos,
- user_pos, user_len, new);
+ user_pos, user_len,
+ should_zero);
if (tmpret) {
mlog_errno(tmpret);
if (ret == 0)
return ret;
}
+static int ocfs2_write_cluster_by_desc(struct address_space *mapping,
+ struct ocfs2_alloc_context *data_ac,
+ struct ocfs2_alloc_context *meta_ac,
+ struct ocfs2_write_ctxt *wc,
+ loff_t pos, unsigned len)
+{
+ int ret, i;
+ loff_t cluster_off;
+ unsigned int local_len = len;
+ struct ocfs2_write_cluster_desc *desc;
+ struct ocfs2_super *osb = OCFS2_SB(mapping->host->i_sb);
+
+ for (i = 0; i < wc->w_clen; i++) {
+ desc = &wc->w_desc[i];
+
+ /*
+ * We have to make sure that the total write passed in
+ * doesn't extend past a single cluster.
+ */
+ local_len = len;
+ cluster_off = pos & (osb->s_clustersize - 1);
+ if ((cluster_off + local_len) > osb->s_clustersize)
+ local_len = osb->s_clustersize - cluster_off;
+
+ ret = ocfs2_write_cluster(mapping, desc->c_phys,
+ desc->c_unwritten, data_ac, meta_ac,
+ wc, desc->c_cpos, pos, local_len);
+ if (ret) {
+ mlog_errno(ret);
+ goto out;
+ }
+
+ len -= local_len;
+ pos += local_len;
+ }
+
+ ret = 0;
+out:
+ return ret;
+}
+
/*
* ocfs2_write_end() wants to know which parts of the target page it
* should complete the write on. It's easiest to compute them ahead of
if (wc->w_large_pages) {
/*
* We only care about the 1st and last cluster within
- * our range and whether they are holes or not. Either
+ * our range and whether they should be zero'd or not. Either
* value may be extended out to the start/end of a
* newly allocated cluster.
*/
desc = &wc->w_desc[0];
- if (desc->c_new)
+ if (ocfs2_should_zero_cluster(desc))
ocfs2_figure_cluster_boundaries(osb,
desc->c_cpos,
&wc->w_target_from,
NULL);
desc = &wc->w_desc[wc->w_clen - 1];
- if (desc->c_new)
+ if (ocfs2_should_zero_cluster(desc))
ocfs2_figure_cluster_boundaries(osb,
desc->c_cpos,
NULL,
}
}
-int ocfs2_write_begin_nolock(struct address_space *mapping,
- loff_t pos, unsigned len, unsigned flags,
- struct page **pagep, void **fsdata,
- struct buffer_head *di_bh, struct page *mmap_page)
+/*
+ * Populate each single-cluster write descriptor in the write context
+ * with information about the i/o to be done.
+ *
+ * Returns the number of clusters that will have to be allocated, as
+ * well as a worst case estimate of the number of extent records that
+ * would have to be created during a write to an unwritten region.
+ */
+static int ocfs2_populate_write_desc(struct inode *inode,
+ struct ocfs2_write_ctxt *wc,
+ unsigned int *clusters_to_alloc,
+ unsigned int *extents_to_split)
{
- int ret, i, credits = OCFS2_INODE_UPDATE_CREDITS;
- unsigned int num_clusters = 0, clusters_to_alloc = 0;
- u32 phys = 0;
- struct ocfs2_write_ctxt *wc;
- struct inode *inode = mapping->host;
- struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
- struct ocfs2_dinode *di;
- struct ocfs2_alloc_context *data_ac = NULL;
- struct ocfs2_alloc_context *meta_ac = NULL;
- handle_t *handle;
+ int ret;
struct ocfs2_write_cluster_desc *desc;
+ unsigned int num_clusters = 0;
+ unsigned int ext_flags = 0;
+ u32 phys = 0;
+ int i;
- ret = ocfs2_alloc_write_ctxt(&wc, osb, pos, len, di_bh);
- if (ret) {
- mlog_errno(ret);
- return ret;
- }
-
- di = (struct ocfs2_dinode *)wc->w_di_bh->b_data;
+ *clusters_to_alloc = 0;
+ *extents_to_split = 0;
for (i = 0; i < wc->w_clen; i++) {
desc = &wc->w_desc[i];
desc->c_cpos = wc->w_cpos + i;
if (num_clusters == 0) {
+ /*
+ * Need to look up the next extent record.
+ */
ret = ocfs2_get_clusters(inode, desc->c_cpos, &phys,
- &num_clusters, NULL);
+ &num_clusters, &ext_flags);
if (ret) {
mlog_errno(ret);
goto out;
}
+
+ /*
+ * Assume worst case - that we're writing in
+ * the middle of the extent.
+ *
+ * We can assume that the write proceeds from
+ * left to right, in which case the extent
+ * insert code is smart enough to coalesce the
+ * next splits into the previous records created.
+ */
+ if (ext_flags & OCFS2_EXT_UNWRITTEN)
+ *extents_to_split = *extents_to_split + 2;
} else if (phys) {
/*
* Only increment phys if it doesn't describe
desc->c_phys = phys;
if (phys == 0) {
desc->c_new = 1;
- clusters_to_alloc++;
+ *clusters_to_alloc = *clusters_to_alloc + 1;
}
+ if (ext_flags & OCFS2_EXT_UNWRITTEN)
+ desc->c_unwritten = 1;
num_clusters--;
}
+ ret = 0;
+out:
+ return ret;
+}
+
+int ocfs2_write_begin_nolock(struct address_space *mapping,
+ loff_t pos, unsigned len, unsigned flags,
+ struct page **pagep, void **fsdata,
+ struct buffer_head *di_bh, struct page *mmap_page)
+{
+ int ret, credits = OCFS2_INODE_UPDATE_CREDITS;
+ unsigned int clusters_to_alloc, extents_to_split;
+ struct ocfs2_write_ctxt *wc;
+ struct inode *inode = mapping->host;
+ struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
+ struct ocfs2_dinode *di;
+ struct ocfs2_alloc_context *data_ac = NULL;
+ struct ocfs2_alloc_context *meta_ac = NULL;
+ handle_t *handle;
+
+ ret = ocfs2_alloc_write_ctxt(&wc, osb, pos, len, di_bh);
+ if (ret) {
+ mlog_errno(ret);
+ return ret;
+ }
+
+ ret = ocfs2_populate_write_desc(inode, wc, &clusters_to_alloc,
+ &extents_to_split);
+ if (ret) {
+ mlog_errno(ret);
+ goto out;
+ }
+
+ di = (struct ocfs2_dinode *)wc->w_di_bh->b_data;
+
/*
* We set w_target_from, w_target_to here so that
* ocfs2_write_end() knows which range in the target page to
* write out. An allocation requires that we write the entire
* cluster range.
*/
- if (clusters_to_alloc > 0) {
+ if (clusters_to_alloc || extents_to_split) {
/*
* XXX: We are stretching the limits of
- * ocfs2_lock_allocators(). It greately over-estimates
+ * ocfs2_lock_allocators(). It greatly over-estimates
* the work to be done.
*/
ret = ocfs2_lock_allocators(inode, di, clusters_to_alloc,
- &data_ac, &meta_ac);
+ extents_to_split, &data_ac, &meta_ac);
if (ret) {
mlog_errno(ret);
goto out;
}
- ocfs2_set_target_boundaries(osb, wc, pos, len, clusters_to_alloc);
+ ocfs2_set_target_boundaries(osb, wc, pos, len,
+ clusters_to_alloc + extents_to_split);
handle = ocfs2_start_trans(osb, credits);
if (IS_ERR(handle)) {
* extent.
*/
ret = ocfs2_grab_pages_for_write(mapping, wc, wc->w_cpos, pos,
- clusters_to_alloc, mmap_page);
+ clusters_to_alloc + extents_to_split,
+ mmap_page);
if (ret) {
mlog_errno(ret);
goto out_commit;
}
- for (i = 0; i < wc->w_clen; i++) {
- desc = &wc->w_desc[i];
-
- ret = ocfs2_write_cluster(mapping, desc->c_phys, data_ac,
- meta_ac, wc, desc->c_cpos, pos, len);
- if (ret) {
- mlog_errno(ret);
- goto out_commit;
- }
+ ret = ocfs2_write_cluster_by_desc(mapping, data_ac, meta_ac, wc, pos,
+ len);
+ if (ret) {
+ mlog_errno(ret);
+ goto out_commit;
}
if (data_ac)
inode->i_mtime = inode->i_ctime = CURRENT_TIME;
di->i_mtime = di->i_ctime = cpu_to_le64(inode->i_mtime.tv_sec);
di->i_mtime_nsec = di->i_ctime_nsec = cpu_to_le32(inode->i_mtime.tv_nsec);
-
ocfs2_journal_dirty(handle, wc->w_di_bh);
ocfs2_commit_trans(osb, handle);
+ ocfs2_run_deallocs(osb, &wc->w_dealloc);
+
ocfs2_free_write_ctxt(wc);
return copied;
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff