alpha/lib/ev6-stxncpy.S

1 /* SPDX-License-Identifier: GPL-2.0 */
3  * arch/alpha/lib/ev6-stxncpy.S
4  * 21264 version contributed by Rick Gorton <rick.gorton@api-networks.com>
6  * Copy no more than COUNT bytes of the null-terminated string from
29  * Furthermore, v0, a3-a5, t11, and $at are untouched.
34  *	ftp.digital.com/pub/Digital/info/semiconductor/literature/dsc-library.html
36  *	E	- either cluster
37  *	U	- upper subcluster; U0 - subcluster U0; U1 - subcluster U1
38  *	L	- lower subcluster; L0 - subcluster L0; L1 - subcluster L1
66 	lda	t2, -1		# E : build a mask against false zero
106 	   the end-of-count bit is set in t8 iff it applies.
113 	negq	t8, t12		# E : find low bit set
133 	ret	(t9)		# L0 : Latency=3
137 	/* Add the end-of-count bit to the eos detection bitmask.  */
140 	br	$a_eos		# L0 : Latency=3
153 	/* Are source and destination co-aligned?  */
161 	srl	a2, 3, a2	# U : a2 = loop counter = (count - 1)/8 (stall)
166 	/* We are co-aligned; take care of a partial first word.  */
182 /* The source and destination are not co-aligned.  Align the destination
194 	   t1 == the shifted low bits of the first source word
195 	   t6 == bytemask that is -1 in dest word bytes */
197 	ldq_u	t2, 8(a1)	# L : Latency=3 load second src word
209 	lda	t6, -1		# E :
224 	extql	t2, a1, t1	# U : position hi-bits of lo word
226 	ldq_u	t2, 8(a1)	# L : read next high-order source word
229 	extqh	t2, a1, t0	# U : position lo-bits of hi word (stall)
242 	   t0 == the shifted low-order bits from the current source word
243 	   t1 == the shifted high-order bits from the previous source word
252 	extql	t2, a1, t1	# U : extract low bits for next time
255 	stq_u	t0, -8(a0)	# U : save the current word
257 	ldq_u	t2, 8(a1)	# U : Latency=3 load high word for next time
260 	extqh	t2, a1, t0	# U : extract low bits (2 cycle stall)
271 	   t0 == the shifted low-order bits from the current source word
272 	   t1 == the shifted high-order bits from the previous source word
280 	stq_u	t0, 0(a0)	# L : the null was in the high-order bits
289 	cmoveq	a2, t6, t8	# E : Latency=2, extra map slot (stall)
296 	negq	t8, t6		# E : isolate low bit set
312 	ret	(t9)		# L0 : Latency=3
314 	  /* Got to end-of-count before end of string.
316 	     t1 == the shifted high-order bits from the previous source word  */
325 	extqh	t2, a1, t0	# U : extract low bits for last word (stall)
331 $u_eocfin:			# end-of-count, final word
333 	br	$u_final	# L0 : Latency=3
349 	lda	t6, -1		# E :
361 	lda	t2, -1		# E : for creating masks later
367 	or	t8, t10, t5	# E : test for end-of-count too
370 	cmoveq	a2, t5, t8	# E : Latency=2, extra map slot
393 	ret	(t9)		# L0 : Latency=3