sun4u-us3/common/memcpy.s

74 	cmp	%o1, %o0	! if from address is >= to use forward copy
76 	sub	%o0, %o1, %o4	! get difference of two addresses
86 	add     %o1, %o2, %o1           ! get to end of source space
96 	ldub	[%o1-1], %o3		! load last byte
98 	sub	%o1, 4, %o1
99 	ldub	[%o1+2], %o3		! load 2nd from last byte
102 	ldub	[%o1+1], %o3		! load 3rd from last byte
105 	ldub	[%o1], %o3		! load 4th from last byte
112 	dec	%o1			! decrement src address
113 	ldub	[%o1], %o3		! read a byte
128 	dec	%o1			! decrement src address
129 	ldub	[%o1], %o3		! read a byte
137 	andcc	%o1, 7, %g0		! chk src long word alignment
147 	prefetch [%o1 - (1 * BLOCK_SIZE)], 20	! into the prefetch cache
150 	prefetch [%o1 - (2 * BLOCK_SIZE)], 20	! into the prefetch cache
154 	prefetch [%o1 - (3 * BLOCK_SIZE)], 20	! into the prefetch cache
156 	prefetch [%o1 - (4 * BLOCK_SIZE)], 20	! into the prefetch cache
158 	prefetch [%o1 - (5 * BLOCK_SIZE)], 20	! into the prefetch cache
159 	ldd	[%o1-8], %d4		! load
162 	ldd	[%o1-16], %d2		! a block of 64 bytes
163 	sub	%o1, 64, %o1		! decrease src ptr by 64
166 	ldd	[%o1+40], %d4
168 	ldd	[%o1+32], %d2
170 	ldd	[%o1+24], %d4
172 	ldd	[%o1+16], %d2
174 	ldd	[%o1+8], %d4
176 	ldd	[%o1], %d2
188 	ldx	[%o1-8], %o4		! load
191 	ldx	[%o1-16], %o3		! a block of 32 bytes
192 	sub	%o1, 32, %o1		! decrease src ptr by 32
194 	ldx	[%o1+8], %o4
197 	ldx	[%o1], %o3
204 	ldx	[%o1-8], %o4		! load and store 16 bytes
205 	sub	%o1, 16, %o1		! decrease src ptr by 16
208 	ldx	[%o1], %o3		!
218 	ldx	[%o1-8], %o4		! load 8 bytes
219 	sub	%o1, 8, %o1		! decrease src ptr by 8
239 	alignaddr %o1, %g0, %o5		! align src
244 	sub	%o1, %o4, %o1		!
326 	ldub	[%o1-1], %o3		! load last byte
328 	sub	%o1, 4, %o1
329 	ldub	[%o1+2], %o3		! load 2nd from last byte
332 	ldub	[%o1+1], %o3		! load 3rd from last byte
335 	ldub	[%o1], %o3		! load 4th from last byte
339 	dec	%o1			! decrement src address
340 	ldub	[%o1], %o3		! read a byte
362 	or	%o0, %o1, %o3		! prepare alignment check
367 	ldub	[%o1], %o3		! read byte
370 	ldub	[%o1+1], %o3		! repeat for a total of 4 bytes
371 	add	%o1, 4, %o1		! advance SRC by 4
373 	ldub	[%o1-2], %o3
376 	ldub	[%o1-1], %o3
385 	ldub	[%o1], %o3		! load one byte
389 	ldub	[%o1+1], %o3		! load second byte
393 	ldub	[%o1+2], %o3		! load third byte
401 	lduw	[%o1], %o3		! read word
405 	add	%o1, 8, %o1		! update SRC
406 	lduw	[%o1-4], %o3		! read word
416 	lduw	[%o1], %o3
417 	add	%o1, 4, %o1
429 	lduw	[%o1], %o3		! read word
434 	ldub	[%o1+4], %o3		! load one byte
437 	ldub	[%o1+5], %o3		! load second byte
441 	ldub	[%o1+6], %o3		! load third byte
449 	neg	%o1, %o3
460 	ldub	[%o1], %o4
462 	inc	%o1
467 	! Now DST is 8-byte aligned.  o0, o1, o2 are current.
470 	andcc	%o1, 0x3, %g0		! test alignment
473 	prefetch [%o1 + (1 * BLOCK_SIZE)], 20
482 	andcc	%o1, 0x7, %g0		! test word alignment
484 	prefetch [%o1 + (2 * BLOCK_SIZE)], 20
491 	prefetch [%o1 + (3 * BLOCK_SIZE)], 20
497 	ld	[%o1], %o4		! load
500 	ld	[%o1+4], %o3		! a block of 16 bytes
501 	add	%o1, 16, %o1		! increase src ptr by 16
503 	ld	[%o1-8], %o4
506 	ld	[%o1-4], %o3
516 	ld	[%o1], %o4		! load 4 bytes
519 	add	%o1, 8, %o1		! increase src ptr by 8
520 	ld	[%o1-4], %o3		! load 4 bytes
529 	ld	[%o1], %o4		! load 4 bytes
531 	add	%o1, 4, %o1		! increase src ptr by 4
539 	ldub	[%o1], %o3		! load one byte
542 	ldub	[%o1+1], %o3		! load second byte
546 	ldub	[%o1+2], %o3		! load third byte
567 	prefetch [%o1 + (3 * BLOCK_SIZE)], 20	! into the l2 cache
573 	ldx	[%o1], %o4		! load
576 	ldx	[%o1+8], %o3		! a block of 32 bytes
577 	add	%o1, 32, %o1		! increase src ptr by 32
579 	ldx	[%o1-16], %o4
582 	ldx	[%o1-8], %o3
589 	ldx	[%o1], %o4		! load and store 16 bytes
590 	add	%o1, 16, %o1		! increase src ptr by 16
593 	ldx	[%o1-8], %o3		!
603 	ldx	[%o1], %o4		! load 8 bytes
604 	add	%o1, 8, %o1		! increase src ptr by 8
617 	prefetch [%o1 + (2 * BLOCK_SIZE)], 21
621 	add	%o1, 8, %o1		! prepare to round SRC upward
650 	!                          o1
653 	!                                   o1
656 	!                                   o1
659 	alignaddr %o1, %g0, %o1		! set GSR.ALIGN and align o1
671 	ldd	[%o1-8], %d0
673 	add	%o1, %o3, %o1		! back up o1
675 	ldda	[%o1]ASI_FL8_P, %d2
676 	inc	%o1
677 	andcc	%o1, 7, %g0
688 	ldd	[%o1], %d2
691 	add	%o1, 8, %o1		! update SRC
695 	ldd	[%o1], %d0
697 	add	%o1, 16, %o1		! update SRC
701 	ldd	[%o1 - 8], %d2
716 	sub	%o1, 8, %o1
721 	! Currently, o1 is pointing to the next double-aligned byte in SRC
722 	! The 8 bytes starting at [o1-8] are available in d0
746 	ldda	[%o1]ASI_FL8_P, %d2
747 	inc	%o1
762 	ldub	[%o1], %o3
765 	inc	%o1
775 	add	%o1, %o3, %o1 		! Back up %o1
779 	ldda	[%o1]ASI_FL8_P, %d2
780 	inc	%o1
790 	ldub	[%o1], %o3
792 	inc	%o1
820 	! %o1 I/O 8-byte aligned (and we've set GSR.ALIGN)
821 	! %d0 I/O already loaded with SRC data from [%o1-8]
828 	! Recall that %o1 is further into SRC than %o0 is into DST
833 	ldd	[%o1], %f2
834 	prefetch [%o1 + (3 * BLOCK_SIZE)], 21
835 	ldd	[%o1 + 0x8], %f4
837 	ldd	[%o1 + 0x10], %f6
839 	ldd	[%o1 + 0x18], %f8
841 	ldd	[%o1 + 0x20], %f10
843 	prefetch [%o1 + (4 * BLOCK_SIZE)], 21
845 	ldd	[%o1 + 0x28], %f12
848 	ldd	[%o1 + 0x30], %f14
850 	ldd	[%o1 + 0x38], %f0
852 	prefetch [%o1 + (5 * BLOCK_SIZE)], 21
853 	add	%o1, BLOCK_SIZE, %o1		! update SRC
858 	ldd	[%o1], %f2
860 	ldd	[%o1 + 0x8], %f4
874 	ldd	[%o1 + 0x10], %f6
876 	ldd	[%o1 + 0x18], %f8
878 	ldd	[%o1 + 0x20], %f10
880 	ldd	[%o1 + 0x28], %f12
882 	ldd	[%o1 + 0x30], %f14
884 	ldd	[%o1 + 0x38], %f0
887 	prefetch [%o1 + (5 * BLOCK_SIZE)], 21
889 	add	%o1, BLOCK_SIZE, %o1	! update SRC
897 	ldd	[%o1], %f2
911 	ldd	[%o1 + 0x8], %f4
912 	ldd	[%o1 + 0x10], %f6
914 	ldd	[%o1 + 0x18], %f8
916 	ldd	[%o1 + 0x20], %f10
918 	ldd	[%o1 + 0x28], %f12
920 	ldd	[%o1 + 0x30], %f14
939 	! %o1 I/O 8-byte aligned (and we've set GSR.ALIGN)
940 	! %d0 I/O already loaded with SRC data from [%o1-8]
947 	! Recall that %o1 is further into SRC than %o0 is into DST
949 	! prefetch [%o1 + (3 * BLOCK_SIZE)], 21
951 	prefetch [%o1 + (4 * BLOCK_SIZE)], 21
952 	prefetch [%o1 + (5 * BLOCK_SIZE)], 21
953 	ldd	[%o1], %f2
954 	prefetch [%o1 + (6 * BLOCK_SIZE)], 21
955 	ldd	[%o1 + 0x8], %f4
957 	ldd	[%o1 + 0x10], %f6
959 	ldd	[%o1 + 0x18], %f8
961 	ldd	[%o1 + 0x20], %f10
964 	ldd	[%o1 + 0x28], %f12
967 	ldd	[%o1 + 0x30], %f14
969 	ldd	[%o1 + 0x38], %f0
971 	prefetch [%o1 + (7 * BLOCK_SIZE)], 21
972 	add	%o1, BLOCK_SIZE, %o1	! update SRC
981 	ldd	[%o1], %f2
983 	ldd	[%o1 + 0x8], %f4
987 	ldd	[%o1 + 0x10], %f6
989 	ldd	[%o1 + 0x18], %f8
991 	ldd	[%o1 + 0x20], %f10
993 	ldd	[%o1 + 0x28], %f12
995 	ldd	[%o1 + 0x30], %f14
997 	ldd	[%o1 + 0x38], %f0
1000 	prefetch [%o1 + (8 * BLOCK_SIZE) + 8], 21
1008 	prefetch [%o1 + (5 * BLOCK_SIZE)], 21
1010 	add	%o1, BLOCK_SIZE, %o1		! update SRC
1019 	ldd	[%o1], %f2
1033 	ldd	[%o1 + 0x8], %f4
1034 	ldd	[%o1 + 0x10], %f6
1036 	ldd	[%o1 + 0x18], %f8
1038 	ldd	[%o1 + 0x20], %f10
1040 	ldd	[%o1 + 0x28], %f12
1042 	ldd	[%o1 + 0x30], %f14