file.c

linux 内核源代码

int ocfs2_getattr(struct vfsmount *mnt,
		  struct dentry *dentry,
		  struct kstat *stat)
{
	struct inode *inode = dentry->d_inode;
	struct super_block *sb = dentry->d_inode->i_sb;
	struct ocfs2_super *osb = sb->s_fs_info;
	int err;

	mlog_entry_void();

	err = ocfs2_inode_revalidate(dentry);
	if (err) {
		if (err != -ENOENT)
			mlog_errno(err);
		goto bail;
	}

	generic_fillattr(inode, stat);

	/* We set the blksize from the cluster size for performance */
	stat->blksize = osb->s_clustersize;

bail:
	mlog_exit(err);

	return err;
}

int ocfs2_permission(struct inode *inode, int mask, struct nameidata *nd)
{
	int ret;

	mlog_entry_void();

	ret = ocfs2_meta_lock(inode, NULL, 0);
	if (ret) {
		if (ret != -ENOENT)
			mlog_errno(ret);
		goto out;
	}

	ret = generic_permission(inode, mask, NULL);

	ocfs2_meta_unlock(inode, 0);
out:
	mlog_exit(ret);
	return ret;
}

static int __ocfs2_write_remove_suid(struct inode *inode,
				     struct buffer_head *bh)
{
	int ret;
	handle_t *handle;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	struct ocfs2_dinode *di;

	mlog_entry("(Inode %llu, mode 0%o)\n",
		   (unsigned long long)OCFS2_I(inode)->ip_blkno, inode->i_mode);

	handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS);
	if (handle == NULL) {
		ret = -ENOMEM;
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_journal_access(handle, inode, bh,
				   OCFS2_JOURNAL_ACCESS_WRITE);
	if (ret < 0) {
		mlog_errno(ret);
		goto out_trans;
	}

	inode->i_mode &= ~S_ISUID;
	if ((inode->i_mode & S_ISGID) && (inode->i_mode & S_IXGRP))
		inode->i_mode &= ~S_ISGID;

	di = (struct ocfs2_dinode *) bh->b_data;
	di->i_mode = cpu_to_le16(inode->i_mode);

	ret = ocfs2_journal_dirty(handle, bh);
	if (ret < 0)
		mlog_errno(ret);

out_trans:
	ocfs2_commit_trans(osb, handle);
out:
	mlog_exit(ret);
	return ret;
}

/*
 * Will look for holes and unwritten extents in the range starting at
 * pos for count bytes (inclusive).
 */
static int ocfs2_check_range_for_holes(struct inode *inode, loff_t pos,
				       size_t count)
{
	int ret = 0;
	unsigned int extent_flags;
	u32 cpos, clusters, extent_len, phys_cpos;
	struct super_block *sb = inode->i_sb;

	cpos = pos >> OCFS2_SB(sb)->s_clustersize_bits;
	clusters = ocfs2_clusters_for_bytes(sb, pos + count) - cpos;

	while (clusters) {
		ret = ocfs2_get_clusters(inode, cpos, &phys_cpos, &extent_len,
					 &extent_flags);
		if (ret < 0) {
			mlog_errno(ret);
			goto out;
		}

		if (phys_cpos == 0 || (extent_flags & OCFS2_EXT_UNWRITTEN)) {
			ret = 1;
			break;
		}

		if (extent_len > clusters)
			extent_len = clusters;

		clusters -= extent_len;
		cpos += extent_len;
	}
out:
	return ret;
}

static int ocfs2_write_remove_suid(struct inode *inode)
{
	int ret;
	struct buffer_head *bh = NULL;
	struct ocfs2_inode_info *oi = OCFS2_I(inode);

	ret = ocfs2_read_block(OCFS2_SB(inode->i_sb),
			       oi->ip_blkno, &bh, OCFS2_BH_CACHED, inode);
	if (ret < 0) {
		mlog_errno(ret);
		goto out;
	}

	ret =  __ocfs2_write_remove_suid(inode, bh);
out:
	brelse(bh);
	return ret;
}

/*
 * Allocate enough extents to cover the region starting at byte offset
 * start for len bytes. Existing extents are skipped, any extents
 * added are marked as "unwritten".
 */
static int ocfs2_allocate_unwritten_extents(struct inode *inode,
					    u64 start, u64 len)
{
	int ret;
	u32 cpos, phys_cpos, clusters, alloc_size;
	u64 end = start + len;
	struct buffer_head *di_bh = NULL;

	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
		ret = ocfs2_read_block(OCFS2_SB(inode->i_sb),
				       OCFS2_I(inode)->ip_blkno, &di_bh,
				       OCFS2_BH_CACHED, inode);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}

		/*
		 * Nothing to do if the requested reservation range
		 * fits within the inode.
		 */
		if (ocfs2_size_fits_inline_data(di_bh, end))
			goto out;

		ret = ocfs2_convert_inline_data_to_extents(inode, di_bh);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}
	}

	/*
	 * We consider both start and len to be inclusive.
	 */
	cpos = start >> OCFS2_SB(inode->i_sb)->s_clustersize_bits;
	clusters = ocfs2_clusters_for_bytes(inode->i_sb, start + len);
	clusters -= cpos;

	while (clusters) {
		ret = ocfs2_get_clusters(inode, cpos, &phys_cpos,
					 &alloc_size, NULL);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}

		/*
		 * Hole or existing extent len can be arbitrary, so
		 * cap it to our own allocation request.
		 */
		if (alloc_size > clusters)
			alloc_size = clusters;

		if (phys_cpos) {
			/*
			 * We already have an allocation at this
			 * region so we can safely skip it.
			 */
			goto next;
		}

		ret = __ocfs2_extend_allocation(inode, cpos, alloc_size, 1);
		if (ret) {
			if (ret != -ENOSPC)
				mlog_errno(ret);
			goto out;
		}

next:
		cpos += alloc_size;
		clusters -= alloc_size;
	}

	ret = 0;
out:

	brelse(di_bh);
	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;
	struct address_space *mapping = inode->i_mapping;

	ocfs2_init_dealloc_ctxt(&dealloc);

	if (byte_len == 0)
		return 0;

	if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) {
		ret = ocfs2_truncate_inline(inode, di_bh, byte_start,
					    byte_start + byte_len, 0);
		if (ret) {
			mlog_errno(ret);
			goto out;
		}
		/*
		 * There's no need to get fancy with the page cache
		 * truncate of an inline-data inode. We're talking
		 * about less than a page here, which will be cached
		 * in the dinode buffer anyway.
		 */
		unmap_mapping_range(mapping, 0, 0, 0);
		truncate_inode_pages(mapping, 0);
		goto out;
	}

	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;
}

/*
 * Parts of this function taken from xfs_change_file_space()
 */
static int __ocfs2_change_file_space(struct file *file, struct inode *inode,
				     loff_t f_pos, unsigned int cmd,
				     struct ocfs2_space_resv *sr,
				     int change_size)
{
	int ret;
	s64 llen;
	loff_t size;
	struct ocfs2_super *osb = OCFS2_SB(inode->i_sb);
	struct buffer_head *di_bh = NULL;
	handle_t *handle;
	unsigned long long max_off = inode->i_sb->s_maxbytes;

	if (ocfs2_is_hard_readonly(osb) || ocfs2_is_soft_readonly(osb))
		return -EROFS;

	mutex_lock(&inode->i_mutex);

	/*
	 * This prevents concurrent writes on other nodes
	 */
	ret = ocfs2_rw_lock(inode, 1);
	if (ret) {
		mlog_errno(ret);
		goto out;
	}

	ret = ocfs2_meta_lock(inode, &di_bh, 1);
	if (ret) {
		mlog_errno(ret);
		goto out_rw_unlock;
	}

	if (inode->i_flags & (S_IMMUTABLE|S_APPEND)) {
		ret = -EPERM;
		goto out_meta_unlock;
	}

	switch (sr->l_whence) {
	case 0: /*SEEK_SET*/
		break;
	case 1: /*SEEK_CUR*/
		sr->l_start += f_pos;
		break;
	case 2: /*SEEK_END*/
		sr->l_start += i_size_read(inode);
		break;
	default:
		ret = -EINVAL;
		goto out_meta_unlock;
	}
	sr->l_whence = 0;

	llen = sr->l_len > 0 ? sr->l_len - 1 : sr->l_len;

	if (sr->l_start < 0
	    || sr->l_start > max_off
	    || (sr->l_start + llen) < 0
	    || (sr->l_start + llen) > max_off) {
		ret = -EINVAL;
		goto out_meta_unlock;
	}
	size = sr->l_start + sr->l_len;

	if (cmd == OCFS2_IOC_RESVSP || cmd == OCFS2_IOC_RESVSP64) {
		if (sr->l_len <= 0) {
			ret = -EINVAL;
			goto out_meta_unlock;
		}