On Wed, Nov 17, 2010 at 02:40:07PM -0800, Nicholas A. Bellinger wrote: > On Wed, 2010-11-17 at 16:38 -0500, Christoph Hellwig wrote: > > The cmd to task mapping is almost the same for all control CDBs, > > except for calling different backend methods to do the backed-specific > > mapping. Merge the code performing the mapping into a single > > new transport_map_control_cmd_to_task helper, and kill the > > transport_map_task callback by simplify calling the backend methods > > directly for the single task for a control CDB, and hardcoding > > ->map_task_SG for data CDBs. > > > > Signed-off-by: Christoph Hellwig <hch@xxxxxx> > > This one looks very reasonable as the extra set of bitwise AND > instructions is limited to control path code. > > As for the first one, I don't really have a strong preference either > way, as long as the difference for extra 4 AND instructions for the bulk > 'fast past' case is really a wash on modern x86_64 silicon.. This one modifies the code touched by the first. There's one if else for data vs control path CDBs and then we switch the different types of control CDBs. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
