[PATCH] IB/ipath: reduce overhead on receive interrupts

Also count the number of interrupts where that works (fastrcvint).  On any
interrupt where the port0 head and tail registers are not equal, just call the
ipath_kreceive code without reading the interrupt status, thus saving the
approximately 0.25usec processor stall waiting for the read to return.  If any
other interrupt bits are set, or head==tail, take the normal path, but that
has been reordered to handle read ahead of pioavail.  Also no longer call
ipath_kreceive() from ipath_qcheck(), because that just seems to make things
worse, and isn't really buying us anything, these days.

Also no longer loop in ipath_kreceive(); better to not hold things off too
long (I saw many cases where we would loop 4-8 times, and handle thousands (up
to 3500) in a single call).

Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Bryan O'Sullivan <bryan.osullivan@qlogic.com>
Cc: "Michael S. Tsirkin" <mst@mellanox.co.il>
Cc: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

drivers/infiniband/hw/ipath/ipath_common.h

@@ -97,8 +97,8 @@ struct infinipath_stats {
 	__u64 sps_hwerrs;
 	/* number of times IB link changed state unexpectedly */
 	__u64 sps_iblink;
-	/* no longer used; left for compatibility */
-	__u64 sps_unused3;
+	/* kernel receive interrupts that didn't read intstat */
+	__u64 sps_fastrcvint;
 	/* number of kernel (port0) packets received */
 	__u64 sps_port0pkts;
 	/* number of "ethernet" packets sent by driver */