* filled.
*/
unsigned c_new;
- unsigned c_unwritten;
+ unsigned c_clear_unwritten;
unsigned c_needs_zero;
};
/* First cluster allocated in a nonsparse extend */
u32 w_first_new_cpos;
+ /* Type of caller. Must be one of buffer, mmap, direct. */
+ ocfs2_write_type_t w_type;
+
struct ocfs2_write_cluster_desc w_desc[OCFS2_MAX_CLUSTERS_PER_PAGE];
/*
static int ocfs2_alloc_write_ctxt(struct ocfs2_write_ctxt **wcp,
struct ocfs2_super *osb, loff_t pos,
- unsigned len, struct buffer_head *di_bh)
+ unsigned len, ocfs2_write_type_t type,
+ struct buffer_head *di_bh)
{
u32 cend;
struct ocfs2_write_ctxt *wc;
wc->w_clen = cend - wc->w_cpos + 1;
get_bh(di_bh);
wc->w_di_bh = di_bh;
+ wc->w_type = type;
if (unlikely(PAGE_CACHE_SHIFT > osb->s_clustersize_bits))
wc->w_large_pages = 1;
to = user_pos + user_len;
struct page *tmppage;
- ocfs2_zero_new_buffers(wc->w_target_page, from, to);
+ if (wc->w_target_page)
+ ocfs2_zero_new_buffers(wc->w_target_page, from, to);
for(i = 0; i < wc->w_num_pages; i++) {
tmppage = wc->w_pages[i];
- if (page_has_buffers(tmppage)) {
+ if (tmppage && page_has_buffers(tmppage)) {
if (ocfs2_should_order_data(inode))
ocfs2_jbd2_file_inode(wc->w_handle, inode);
wc->w_num_pages = 1;
start = target_index;
}
+ end_index = (user_pos + user_len - 1) >> PAGE_CACHE_SHIFT;
for(i = 0; i < wc->w_num_pages; i++) {
index = start + i;
- if (index == target_index && mmap_page) {
+ if (index >= target_index && index <= end_index &&
+ wc->w_type == OCFS2_WRITE_MMAP) {
/*
* ocfs2_pagemkwrite() is a little different
* and wants us to directly use the page
page_cache_get(mmap_page);
wc->w_pages[i] = mmap_page;
wc->w_target_locked = true;
+ } else if (index >= target_index && index <= end_index &&
+ wc->w_type == OCFS2_WRITE_DIRECT) {
+ /* Direct write has no mapping page. */
+ wc->w_pages[i] = NULL;
+ continue;
} else {
wc->w_pages[i] = find_or_create_page(mapping, index,
GFP_NOFS);
* Prepare a single cluster for write one cluster into the file.
*/
static int ocfs2_write_cluster(struct address_space *mapping,
- u32 phys, unsigned int unwritten,
+ u32 *phys, unsigned int new,
+ unsigned int clear_unwritten,
unsigned int should_zero,
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;
- u64 v_blkno, p_blkno;
+ int ret, i;
+ u64 p_blkno;
struct inode *inode = mapping->host;
struct ocfs2_extent_tree et;
+ int bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1);
- new = phys == 0 ? 1 : 0;
if (new) {
u32 tmp_pos;
*/
tmp_pos = cpos;
ret = ocfs2_add_inode_data(OCFS2_SB(inode->i_sb), inode,
- &tmp_pos, 1, 0, wc->w_di_bh,
- wc->w_handle, data_ac,
- meta_ac, NULL);
+ &tmp_pos, 1, !clear_unwritten,
+ wc->w_di_bh, wc->w_handle,
+ data_ac, meta_ac, NULL);
/*
* This shouldn't happen because we must have already
* calculated the correct meta data allocation required. The
mlog_errno(ret);
goto out;
}
- } else if (unwritten) {
+ } else if (clear_unwritten) {
ocfs2_init_dinode_extent_tree(&et, INODE_CACHE(inode),
wc->w_di_bh);
ret = ocfs2_mark_extent_written(inode, &et,
- wc->w_handle, cpos, 1, phys,
+ wc->w_handle, cpos, 1, *phys,
meta_ac, &wc->w_dealloc);
if (ret < 0) {
mlog_errno(ret);
}
}
- if (should_zero)
- v_blkno = ocfs2_clusters_to_blocks(inode->i_sb, cpos);
- else
- v_blkno = user_pos >> inode->i_sb->s_blocksize_bits;
-
/*
* The only reason this should fail is due to an inability to
* find the extent added.
*/
- ret = ocfs2_extent_map_get_blocks(inode, v_blkno, &p_blkno, NULL,
- NULL);
+ ret = ocfs2_get_clusters(inode, cpos, phys, NULL, NULL);
if (ret < 0) {
mlog(ML_ERROR, "Get physical blkno failed for inode %llu, "
- "at logical block %llu",
- (unsigned long long)OCFS2_I(inode)->ip_blkno,
- (unsigned long long)v_blkno);
+ "at logical cluster %u",
+ (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos);
goto out;
}
- BUG_ON(p_blkno == 0);
+ BUG_ON(*phys == 0);
+
+ p_blkno = ocfs2_clusters_to_blocks(inode->i_sb, *phys);
+ if (!should_zero)
+ p_blkno += (user_pos >> inode->i_sb->s_blocksize_bits) & (u64)(bpc - 1);
for(i = 0; i < wc->w_num_pages; i++) {
int tmpret;
+ /* This is the direct io target page. */
+ if (wc->w_pages[i] == NULL) {
+ p_blkno++;
+ continue;
+ }
+
tmpret = ocfs2_prepare_page_for_write(inode, &p_blkno, wc,
wc->w_pages[i], cpos,
user_pos, user_len,
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,
+ ret = ocfs2_write_cluster(mapping, &desc->c_phys,
+ desc->c_new,
+ desc->c_clear_unwritten,
desc->c_needs_zero,
data_ac, meta_ac,
wc, desc->c_cpos, pos, local_len);
if (phys == 0) {
desc->c_new = 1;
desc->c_needs_zero = 1;
+ desc->c_clear_unwritten = 1;
*clusters_to_alloc = *clusters_to_alloc + 1;
}
if (ext_flags & OCFS2_EXT_UNWRITTEN) {
- desc->c_unwritten = 1;
+ desc->c_clear_unwritten = 1;
desc->c_needs_zero = 1;
}
if (ret)
mlog_errno(ret);
- wc->w_first_new_cpos =
- ocfs2_clusters_for_bytes(inode->i_sb, i_size_read(inode));
+ /* There is no wc if this is call from direct. */
+ if (wc)
+ wc->w_first_new_cpos =
+ ocfs2_clusters_for_bytes(inode->i_sb, i_size_read(inode));
return ret;
}
return ret;
}
-int ocfs2_write_begin_nolock(struct file *filp,
- struct address_space *mapping,
- loff_t pos, unsigned len, unsigned flags,
+int ocfs2_write_begin_nolock(struct address_space *mapping,
+ loff_t pos, unsigned len, ocfs2_write_type_t type,
struct page **pagep, void **fsdata,
struct buffer_head *di_bh, struct page *mmap_page)
{
int try_free = 1, ret1;
try_again:
- ret = ocfs2_alloc_write_ctxt(&wc, osb, pos, len, di_bh);
+ ret = ocfs2_alloc_write_ctxt(&wc, osb, pos, len, type, di_bh);
if (ret) {
mlog_errno(ret);
return ret;
}
}
- if (ocfs2_sparse_alloc(osb))
- ret = ocfs2_zero_tail(inode, di_bh, pos);
- else
- ret = ocfs2_expand_nonsparse_inode(inode, di_bh, pos, len,
- wc);
- if (ret) {
- mlog_errno(ret);
- goto out;
+ /* Direct io change i_size late, should not zero tail here. */
+ if (type != OCFS2_WRITE_DIRECT) {
+ if (ocfs2_sparse_alloc(osb))
+ ret = ocfs2_zero_tail(inode, di_bh, pos);
+ else
+ ret = ocfs2_expand_nonsparse_inode(inode, di_bh, pos,
+ len, wc);
+ if (ret) {
+ mlog_errno(ret);
+ goto out;
+ }
}
ret = ocfs2_check_range_for_refcount(inode, pos, len);
(unsigned long long)OCFS2_I(inode)->ip_blkno,
(long long)i_size_read(inode),
le32_to_cpu(di->i_clusters),
- pos, len, flags, mmap_page,
+ pos, len, type, mmap_page,
clusters_to_alloc, extents_to_split);
/*
credits = ocfs2_calc_extend_credits(inode->i_sb,
&di->id2.i_list);
-
- }
+ } else if (type == OCFS2_WRITE_DIRECT)
+ /* direct write needs not to start trans if no extents alloc. */
+ goto success;
/*
* We have to zero sparse allocated clusters, unwritten extent clusters,
ocfs2_free_alloc_context(meta_ac);
success:
- *pagep = wc->w_target_page;
+ if (pagep)
+ *pagep = wc->w_target_page;
*fsdata = wc;
return 0;
out_quota:
*/
down_write(&OCFS2_I(inode)->ip_alloc_sem);
- ret = ocfs2_write_begin_nolock(file, mapping, pos, len, flags, pagep,
- fsdata, di_bh, NULL);
+ ret = ocfs2_write_begin_nolock(mapping, pos, len, OCFS2_WRITE_BUFFER,
+ pagep, fsdata, di_bh, NULL);
if (ret) {
mlog_errno(ret);
goto out_fail;
handle_t *handle = wc->w_handle;
struct page *tmppage;
- ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), wc->w_di_bh,
- OCFS2_JOURNAL_ACCESS_WRITE);
- if (ret) {
- copied = ret;
- mlog_errno(ret);
- goto out;
+ if (handle) {
+ ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode),
+ wc->w_di_bh, OCFS2_JOURNAL_ACCESS_WRITE);
+ if (ret) {
+ copied = ret;
+ mlog_errno(ret);
+ goto out;
+ }
}
if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
goto out_write_size;
}
- if (unlikely(copied < len)) {
+ if (unlikely(copied < len) && wc->w_target_page) {
if (!PageUptodate(wc->w_target_page))
copied = 0;
ocfs2_zero_new_buffers(wc->w_target_page, start+copied,
start+len);
}
- flush_dcache_page(wc->w_target_page);
+ if (wc->w_target_page)
+ flush_dcache_page(wc->w_target_page);
for(i = 0; i < wc->w_num_pages; i++) {
tmppage = wc->w_pages[i];
+ /* This is the direct io target page. */
+ if (tmppage == NULL)
+ continue;
+
if (tmppage == wc->w_target_page) {
from = wc->w_target_from;
to = wc->w_target_to;
}
if (page_has_buffers(tmppage)) {
- if (ocfs2_should_order_data(inode))
- ocfs2_jbd2_file_inode(wc->w_handle, inode);
+ if (handle && ocfs2_should_order_data(inode))
+ ocfs2_jbd2_file_inode(handle, inode);
block_commit_write(tmppage, from, to);
}
}
out_write_size:
- pos += copied;
- if (pos > i_size_read(inode)) {
- i_size_write(inode, pos);
- mark_inode_dirty(inode);
- }
- inode->i_blocks = ocfs2_inode_sector_count(inode);
- di->i_size = cpu_to_le64((u64)i_size_read(inode));
- 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_update_inode_fsync_trans(handle, inode, 1);
- ocfs2_journal_dirty(handle, wc->w_di_bh);
+ /* Direct io do not update i_size here. */
+ if (wc->w_type != OCFS2_WRITE_DIRECT) {
+ pos += copied;
+ if (pos > i_size_read(inode)) {
+ i_size_write(inode, pos);
+ mark_inode_dirty(inode);
+ }
+ inode->i_blocks = ocfs2_inode_sector_count(inode);
+ di->i_size = cpu_to_le64((u64)i_size_read(inode));
+ 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_update_inode_fsync_trans(handle, inode, 1);
+ }
+ if (handle)
+ ocfs2_journal_dirty(handle, wc->w_di_bh);
out:
/* unlock pages before dealloc since it needs acquiring j_trans_barrier
*/
ocfs2_unlock_pages(wc);
- ocfs2_commit_trans(osb, handle);
+ if (handle)
+ ocfs2_commit_trans(osb, handle);
ocfs2_run_deallocs(osb, &wc->w_dealloc);