ATA rman performance enhancement

While doing some benchmarking of other code, I noticed that there were a 
lot of calls to rman_get_bustag/handle().  They weren't taking up much 
actual time since they're pretty lightweight but seemed to be unnecessary.

I worked up the attached diff and benchmarked it.  There are about 1000 
calls a second to the rman routines without this patch and essentially 
none with it.  It makes about a 1% difference in throughput under some 
IO loads.  It is only for non-Promise or non-SII controllers right now 
since I didn't extend the initialization step to more than ata-pci.c. 
The same approach could be used for the other INW/OUTW calls as well but 
they're not in the fast path.  I think it may make more of a difference 
with small reads.

Feel free to test, cleanup, and commit.

Thanks,
-- 
Nate

Index: ata-all.h
===================================================================
RCS file: /home/ncvs/src/sys/dev/ata/ata-all.h,v
retrieving revision 1.87
diff -u -r1.87 ata-all.h
--- ata-all.h	9 Dec 2004 07:31:06 -0000	1.87
+++ ata-all.h	3 Jan 2005 09:54:40 -0000
_at__at_ -331,6 +331,8 _at__at_
 struct ata_resource {
     struct resource		*res;
     int				offset;
+    bus_space_tag_t		tag;
+    bus_space_handle_t		handle;
 };
 
 /* structure describing an ATA channel */
_at__at_ -493,7 +495,7 _at__at_
 				       (offset), (addr), (count))
 
 #define ATA_IDX_INB(ch, idx) \
-	ATA_INB(ch->r_io[idx].res, ch->r_io[idx].offset)
+	bus_space_read_1(ch->r_io[idx].tag, ch->r_io[idx].handle, ch->r_io[idx].offset)
 
 #define ATA_IDX_INW(ch, idx) \
 	ATA_INW(ch->r_io[idx].res, ch->r_io[idx].offset)
_at__at_ -514,13 +516,13 _at__at_
 	ATA_INSL_STRM(ch->r_io[idx].res, ch->r_io[idx].offset, addr, count)
 
 #define ATA_IDX_OUTB(ch, idx, value) \
-	ATA_OUTB(ch->r_io[idx].res, ch->r_io[idx].offset, value)
+	bus_space_write_1(ch->r_io[idx].tag, ch->r_io[idx].handle, ch->r_io[idx].offset, value)
 
 #define ATA_IDX_OUTW(ch, idx, value) \
 	ATA_OUTW(ch->r_io[idx].res, ch->r_io[idx].offset, value)
 
 #define ATA_IDX_OUTL(ch, idx, value) \
-	ATA_OUTL(ch->r_io[idx].res, ch->r_io[idx].offset, value)
+	bus_space_write_4(ch->r_io[idx].tag, ch->r_io[idx].handle, ch->r_io[idx].offset, value)
 
 #define ATA_IDX_OUTSW(ch, idx, addr, count) \
 	ATA_OUTSW(ch->r_io[idx].res, ch->r_io[idx].offset, addr, count)
Index: ata-pci.c
===================================================================
RCS file: /home/ncvs/src/sys/dev/ata/ata-pci.c,v
retrieving revision 1.91
diff -u -r1.91 ata-pci.c
--- ata-pci.c	8 Dec 2004 11:17:38 -0000	1.91
+++ ata-pci.c	3 Jan 2005 09:45:55 -0000
_at__at_ -417,15 +417,23 _at__at_
 
     for (i = ATA_DATA; i <= ATA_STATUS; i ++) {
 	ch->r_io[i].res = io;
+	ch->r_io[i].tag = rman_get_bustag(io);
+	ch->r_io[i].handle = rman_get_bushandle(io);
 	ch->r_io[i].offset = i;
     }
     ch->r_io[ATA_ALTSTAT].res = altio;
+    ch->r_io[ATA_ALTSTAT].tag = rman_get_bustag(altio);
+    ch->r_io[ATA_ALTSTAT].handle = rman_get_bushandle(altio);
     ch->r_io[ATA_ALTSTAT].offset = ata_legacy(device_get_parent(dev)) ? 0 : 2;
     ch->r_io[ATA_IDX_ADDR].res = io;
+    ch->r_io[ATA_IDX_ADDR].tag = rman_get_bustag(io);
+    ch->r_io[ATA_IDX_ADDR].handle = rman_get_bushandle(io);
 
     if (ctlr->r_res1) {
 	for (i = ATA_BMCMD_PORT; i <= ATA_BMDTP_PORT; i++) {
 	    ch->r_io[i].res = ctlr->r_res1;
+	    ch->r_io[i].tag = rman_get_bustag(ctlr->r_res1);
+	    ch->r_io[i].handle = rman_get_bushandle(ctlr->r_res1);
 	    ch->r_io[i].offset = (i - ATA_BMCMD_PORT)+(ch->unit * ATA_BMIOSIZE);
 	}
     }