Commit 55f2a045 authored by Ilya Dryomov's avatar Ilya Dryomov
Browse files

ceph: remove special ack vs commit behavior 



- ask for a commit reply instead of an ack reply in
  __ceph_pool_perm_get()
- don't ask for both ack and commit replies in ceph_sync_write()
- since just only one reply is requested now, i_unsafe_writes list
  will always be empty -- kill ceph_sync_write_wait() and go back to
  a standard ->evict_inode()

Signed-off-by: default avatarIlya Dryomov <idryomov@gmail.com>
Reviewed-by: default avatarJeff Layton <jlayton@redhat.com>
Reviewed-by: default avatarSage Weil <sage@redhat.com>
parent f1075480

fs/ceph/addr.c

+1 −1
Original line number Diff line number Diff line
@@ -1872,7 +1872,7 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, 
		goto out_unlock;

	}



	wr_req->r_flags = CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ACK;

	wr_req->r_flags = CEPH_OSD_FLAG_WRITE | CEPH_OSD_FLAG_ONDISK;

	osd_req_op_init(wr_req, 0, CEPH_OSD_OP_CREATE, CEPH_OSD_OP_FLAG_EXCL);

	ceph_oloc_copy(&wr_req->r_base_oloc, &rd_req->r_base_oloc);

	ceph_oid_copy(&wr_req->r_base_oid, &rd_req->r_base_oid);

fs/ceph/caps.c

+0 −2
Original line number Diff line number Diff line
@@ -2091,8 +2091,6 @@ int ceph_fsync(struct file *file, loff_t start, loff_t end, int datasync) 


	dout("fsync %p%s\n", inode, datasync ? " datasync" : "");



	ceph_sync_write_wait(inode);



	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);

	if (ret < 0)

		goto out;

fs/ceph/file.c

+1 −87
Original line number Diff line number Diff line
@@ -795,89 +795,6 @@ out: 
	kfree(aio_work);

}



/*

 * Write commit request unsafe callback, called to tell us when a

 * request is unsafe (that is, in flight--has been handed to the

 * messenger to send to its target osd).  It is called again when

 * we've received a response message indicating the request is

 * "safe" (its CEPH_OSD_FLAG_ONDISK flag is set), or when a request

 * is completed early (and unsuccessfully) due to a timeout or

 * interrupt.

 *

 * This is used if we requested both an ACK and ONDISK commit reply

 * from the OSD.

 */

static void ceph_sync_write_unsafe(struct ceph_osd_request *req, bool unsafe)

{

	struct ceph_inode_info *ci = ceph_inode(req->r_inode);



	dout("%s %p tid %llu %ssafe\n", __func__, req, req->r_tid,

		unsafe ? "un" : "");

	if (unsafe) {

		ceph_get_cap_refs(ci, CEPH_CAP_FILE_WR);

		spin_lock(&ci->i_unsafe_lock);

		list_add_tail(&req->r_unsafe_item,

			      &ci->i_unsafe_writes);

		spin_unlock(&ci->i_unsafe_lock);



		complete_all(&req->r_completion);

	} else {

		spin_lock(&ci->i_unsafe_lock);

		list_del_init(&req->r_unsafe_item);

		spin_unlock(&ci->i_unsafe_lock);

		ceph_put_cap_refs(ci, CEPH_CAP_FILE_WR);

	}

}



/*

 * Wait on any unsafe replies for the given inode.  First wait on the

 * newest request, and make that the upper bound.  Then, if there are

 * more requests, keep waiting on the oldest as long as it is still older

 * than the original request.

 */

void ceph_sync_write_wait(struct inode *inode)

{

	struct ceph_inode_info *ci = ceph_inode(inode);

	struct list_head *head = &ci->i_unsafe_writes;

	struct ceph_osd_request *req;

	u64 last_tid;



	if (!S_ISREG(inode->i_mode))

		return;



	spin_lock(&ci->i_unsafe_lock);

	if (list_empty(head))

		goto out;



	/* set upper bound as _last_ entry in chain */



	req = list_last_entry(head, struct ceph_osd_request,

			      r_unsafe_item);

	last_tid = req->r_tid;



	do {

		ceph_osdc_get_request(req);

		spin_unlock(&ci->i_unsafe_lock);



		dout("sync_write_wait on tid %llu (until %llu)\n",

		     req->r_tid, last_tid);

		wait_for_completion(&req->r_done_completion);

		ceph_osdc_put_request(req);



		spin_lock(&ci->i_unsafe_lock);

		/*

		 * from here on look at first entry in chain, since we

		 * only want to wait for anything older than last_tid

		 */

		if (list_empty(head))

			break;

		req = list_first_entry(head, struct ceph_osd_request,

				       r_unsafe_item);

	} while (req->r_tid < last_tid);

out:

	spin_unlock(&ci->i_unsafe_lock);

}



static ssize_t

ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter,

		       struct ceph_snap_context *snapc,
@@ -1119,8 +1036,7 @@ ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos, 


	flags = CEPH_OSD_FLAG_ORDERSNAP |

		CEPH_OSD_FLAG_ONDISK |

		CEPH_OSD_FLAG_WRITE |

		CEPH_OSD_FLAG_ACK;

		CEPH_OSD_FLAG_WRITE;



	while ((len = iov_iter_count(from)) > 0) {

		size_t left;
@@ -1166,8 +1082,6 @@ ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos, 
			goto out;

		}



		/* get a second commit callback */

		req->r_unsafe_callback = ceph_sync_write_unsafe;

		req->r_inode = inode;



		osd_req_op_extent_osd_data_pages(req, 0, pages, len, 0,

fs/ceph/inode.c

+0 −9
Original line number Diff line number Diff line
@@ -499,7 +499,6 @@ struct inode *ceph_alloc_inode(struct super_block *sb) 
	ci->i_rdcache_gen = 0;

	ci->i_rdcache_revoking = 0;



	INIT_LIST_HEAD(&ci->i_unsafe_writes);

	INIT_LIST_HEAD(&ci->i_unsafe_dirops);

	INIT_LIST_HEAD(&ci->i_unsafe_iops);

	spin_lock_init(&ci->i_unsafe_lock);
@@ -583,14 +582,6 @@ int ceph_drop_inode(struct inode *inode) 
	return 1;

}



void ceph_evict_inode(struct inode *inode)

{

	/* wait unsafe sync writes */

	ceph_sync_write_wait(inode);

	truncate_inode_pages_final(&inode->i_data);

	clear_inode(inode);

}



static inline blkcnt_t calc_inode_blocks(u64 size)

{

	return (size + (1<<9) - 1) >> 9;

fs/ceph/super.c

+0 −1
Original line number Diff line number Diff line
@@ -757,7 +757,6 @@ static const struct super_operations ceph_super_ops = { 
	.destroy_inode	= ceph_destroy_inode,

	.write_inode    = ceph_write_inode,

	.drop_inode	= ceph_drop_inode,

	.evict_inode	= ceph_evict_inode,

	.sync_fs        = ceph_sync_fs,

	.put_super	= ceph_put_super,

	.show_options   = ceph_show_options,

CRA Git | Maintained and supported by SUSTech CRA and CCSE