[SCSI] libosd: Add Flush and List-objects support
authorBoaz Harrosh <bharrosh@panasas.com>
Sun, 25 Jan 2009 15:05:07 +0000 (17:05 +0200)
committerJames Bottomley <James.Bottomley@HansenPartnership.com>
Thu, 12 Mar 2009 17:58:07 +0000 (12:58 -0500)
Add support for the various List-objects commands. List-partitions-in-device,
List-collections-in-partition, List-objects-in-partition,
List-objects-in-collection. All these support partial listing and continuation.

Add support for the different Flush commands and options.

Signed-off-by: Boaz Harrosh <bharrosh@panasas.com>
Reviewed-by: Benny Halevy <bhalevy@panasas.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
drivers/scsi/osd/osd_initiator.c

index 6849f26..5d9457b 100644 (file)
@@ -104,6 +104,16 @@ static bool _osd_req_is_alist_type(struct osd_request *or,
        }
 }
 
        }
 }
 
+/* This is for List-objects not Attributes-Lists */
+static void _osd_req_encode_olist(struct osd_request *or,
+       struct osd_obj_id_list *list)
+{
+       struct osd_cdb_head *cdbh = osd_cdb_head(&or->cdb);
+
+       cdbh->v1.list_identifier = list->list_identifier;
+       cdbh->v1.start_address = list->continuation_id;
+}
+
 static osd_cdb_offset osd_req_encode_offset(struct osd_request *or,
        u64 offset, unsigned *padding)
 {
 static osd_cdb_offset osd_req_encode_offset(struct osd_request *or,
        u64 offset, unsigned *padding)
 {
@@ -340,6 +350,29 @@ void osd_req_format(struct osd_request *or, u64 tot_capacity)
 }
 EXPORT_SYMBOL(osd_req_format);
 
 }
 EXPORT_SYMBOL(osd_req_format);
 
+int osd_req_list_dev_partitions(struct osd_request *or,
+       osd_id initial_id, struct osd_obj_id_list *list, unsigned nelem)
+{
+       return osd_req_list_partition_objects(or, 0, initial_id, list, nelem);
+}
+EXPORT_SYMBOL(osd_req_list_dev_partitions);
+
+static void _osd_req_encode_flush(struct osd_request *or,
+       enum osd_options_flush_scope_values op)
+{
+       struct osd_cdb_head *ocdb = osd_cdb_head(&or->cdb);
+
+       ocdb->command_specific_options = op;
+}
+
+void osd_req_flush_obsd(struct osd_request *or,
+       enum osd_options_flush_scope_values op)
+{
+       _osd_req_encode_common(or, OSD_ACT_FLUSH_OSD, &osd_root_object, 0, 0);
+       _osd_req_encode_flush(or, op);
+}
+EXPORT_SYMBOL(osd_req_flush_obsd);
+
 /*
  * Partition commands
  */
 /*
  * Partition commands
  */
@@ -366,6 +399,88 @@ void osd_req_remove_partition(struct osd_request *or, osd_id partition)
 }
 EXPORT_SYMBOL(osd_req_remove_partition);
 
 }
 EXPORT_SYMBOL(osd_req_remove_partition);
 
+static int _osd_req_list_objects(struct osd_request *or,
+       __be16 action, const struct osd_obj_id *obj, osd_id initial_id,
+       struct osd_obj_id_list *list, unsigned nelem)
+{
+       struct request_queue *q = or->osd_dev->scsi_device->request_queue;
+       u64 len = nelem * sizeof(osd_id) + sizeof(*list);
+       struct bio *bio;
+
+       _osd_req_encode_common(or, action, obj, (u64)initial_id, len);
+
+       if (list->list_identifier)
+               _osd_req_encode_olist(or, list);
+
+       WARN_ON(or->in.bio);
+       bio = bio_map_kern(q, list, len, or->alloc_flags);
+       if (!bio) {
+               OSD_ERR("!!! Failed to allocate list_objects BIO\n");
+               return -ENOMEM;
+       }
+
+       bio->bi_rw &= ~(1 << BIO_RW);
+       or->in.bio = bio;
+       or->in.total_bytes = bio->bi_size;
+       return 0;
+}
+
+int osd_req_list_partition_collections(struct osd_request *or,
+       osd_id partition, osd_id initial_id, struct osd_obj_id_list *list,
+       unsigned nelem)
+{
+       struct osd_obj_id par = {
+               .partition = partition,
+               .id = 0,
+       };
+
+       return osd_req_list_collection_objects(or, &par, initial_id, list,
+                                              nelem);
+}
+EXPORT_SYMBOL(osd_req_list_partition_collections);
+
+int osd_req_list_partition_objects(struct osd_request *or,
+       osd_id partition, osd_id initial_id, struct osd_obj_id_list *list,
+       unsigned nelem)
+{
+       struct osd_obj_id par = {
+               .partition = partition,
+               .id = 0,
+       };
+
+       return _osd_req_list_objects(or, OSD_ACT_LIST, &par, initial_id, list,
+                                    nelem);
+}
+EXPORT_SYMBOL(osd_req_list_partition_objects);
+
+void osd_req_flush_partition(struct osd_request *or,
+       osd_id partition, enum osd_options_flush_scope_values op)
+{
+       _osd_req_encode_partition(or, OSD_ACT_FLUSH_PARTITION, partition);
+       _osd_req_encode_flush(or, op);
+}
+EXPORT_SYMBOL(osd_req_flush_partition);
+
+/*
+ * Collection commands
+ */
+int osd_req_list_collection_objects(struct osd_request *or,
+       const struct osd_obj_id *obj, osd_id initial_id,
+       struct osd_obj_id_list *list, unsigned nelem)
+{
+       return _osd_req_list_objects(or, OSD_ACT_LIST_COLLECTION, obj,
+                                    initial_id, list, nelem);
+}
+EXPORT_SYMBOL(osd_req_list_collection_objects);
+
+void osd_req_flush_collection(struct osd_request *or,
+       const struct osd_obj_id *obj, enum osd_options_flush_scope_values op)
+{
+       _osd_req_encode_common(or, OSD_ACT_FLUSH_PARTITION, obj, 0, 0);
+       _osd_req_encode_flush(or, op);
+}
+EXPORT_SYMBOL(osd_req_flush_collection);
+
 /*
  * Object commands
  */
 /*
  * Object commands
  */
@@ -392,6 +507,15 @@ void osd_req_write(struct osd_request *or,
 }
 EXPORT_SYMBOL(osd_req_write);
 
 }
 EXPORT_SYMBOL(osd_req_write);
 
+void osd_req_flush_object(struct osd_request *or,
+       const struct osd_obj_id *obj, enum osd_options_flush_scope_values op,
+       /*V2*/ u64 offset, /*V2*/ u64 len)
+{
+       _osd_req_encode_common(or, OSD_ACT_FLUSH, obj, offset, len);
+       _osd_req_encode_flush(or, op);
+}
+EXPORT_SYMBOL(osd_req_flush_object);
+
 void osd_req_read(struct osd_request *or,
        const struct osd_obj_id *obj, struct bio *bio, u64 offset)
 {
 void osd_req_read(struct osd_request *or,
        const struct osd_obj_id *obj, struct bio *bio, u64 offset)
 {