[SCSI] update SG_ALL to avoid causing chaining

Since the sg chaining patches went in, our current value of 255 for
SG_ALL excites chaining on some drivers which cannot support it (and
would thus oops).  Redefine SG_ALL to mean no sg table size
preference, but use the single allocation (non chained) limit.  This
also helps for drivers that use it to size an internal table.

We'll do an opt in system later where truly chaining supporting
drivers can define their sg_tablesize to be anything up to
SCSI_MAX_SG_CHAIN_ELEMENTS.
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>

include/scsi/scsi_host.h

@@ -6,6 +6,7 @@
 #include <linux/types.h>
 #include <linux/workqueue.h>
 #include <linux/mutex.h>
+#include <scsi/scsi.h>
 
 struct request_queue;
 struct block_device;
@@ -25,12 +26,15 @@ struct blk_queue_tags;
  * NONE: Self evident.	Host adapter is not capable of scatter-gather.
  * ALL:	 Means that the host adapter module can do scatter-gather,
  *	 and that there is no limit to the size of the table to which
- *	 we scatter/gather data.
+ *	 we scatter/gather data.  The value we set here is the maximum
+ *	 single element sglist.  To use chained sglists, the adapter
+ *	 has to set a value beyond ALL (and correctly use the chain
+ *	 handling API.
  * Anything else:  Indicates the maximum number of chains that can be
  *	 used in one scatter-gather request.
  */
 #define SG_NONE 0
-#define SG_ALL 0xff
+#define SG_ALL	SCSI_MAX_SG_SEGMENTS
 
 #define MODE_UNKNOWN 0x00
 #define MODE_INITIATOR 0x01