Text file
src/crypto/sha256/sha256block_ppc64x.s
1// Copyright 2016 The Go Authors. All rights reserved.
2// Use of this source code is governed by a BSD-style
3// license that can be found in the LICENSE file.
4
5//go:build ppc64 || ppc64le
6
7// Based on CRYPTOGAMS code with the following comment:
8// # ====================================================================
9// # Written by Andy Polyakov <appro@openssl.org> for the OpenSSL
10// # project. The module is, however, dual licensed under OpenSSL and
11// # CRYPTOGAMS licenses depending on where you obtain it. For further
12// # details see http://www.openssl.org/~appro/cryptogams/.
13// # ====================================================================
14
15#include "textflag.h"
16
17// SHA256 block routine. See sha256block.go for Go equivalent.
18//
19// The algorithm is detailed in FIPS 180-4:
20//
21// https://csrc.nist.gov/publications/fips/fips180-4/fips-180-4.pdf
22//
23// Wt = Mt; for 0 <= t <= 15
24// Wt = SIGMA1(Wt-2) + SIGMA0(Wt-15) + Wt-16; for 16 <= t <= 63
25//
26// a = H0
27// b = H1
28// c = H2
29// d = H3
30// e = H4
31// f = H5
32// g = H6
33// h = H7
34//
35// for t = 0 to 63 {
36// T1 = h + BIGSIGMA1(e) + Ch(e,f,g) + Kt + Wt
37// T2 = BIGSIGMA0(a) + Maj(a,b,c)
38// h = g
39// g = f
40// f = e
41// e = d + T1
42// d = c
43// c = b
44// b = a
45// a = T1 + T2
46// }
47//
48// H0 = a + H0
49// H1 = b + H1
50// H2 = c + H2
51// H3 = d + H3
52// H4 = e + H4
53// H5 = f + H5
54// H6 = g + H6
55// H7 = h + H7
56
57#define CTX R3
58#define INP R4
59#define END R5
60#define TBL R6 // Pointer into kcon table
61#define LEN R9
62#define TEMP R12
63
64#define TBL_STRT R7 // Pointer to start of kcon table.
65
66#define R_x000 R0
67#define R_x010 R8
68#define R_x020 R10
69#define R_x030 R11
70#define R_x040 R14
71#define R_x050 R15
72#define R_x060 R16
73#define R_x070 R17
74#define R_x080 R18
75#define R_x090 R19
76#define R_x0a0 R20
77#define R_x0b0 R21
78#define R_x0c0 R22
79#define R_x0d0 R23
80#define R_x0e0 R24
81#define R_x0f0 R25
82#define R_x100 R26
83#define R_x110 R27
84
85
86// V0-V7 are A-H
87// V8-V23 are used for the message schedule
88#define KI V24
89#define FUNC V25
90#define S0 V26
91#define S1 V27
92#define s0 V28
93#define s1 V29
94#define LEMASK V31 // Permutation control register for little endian
95
96// 4 copies of each Kt, to fill all 4 words of a vector register
97DATA ·kcon+0x000(SB)/8, $0x428a2f98428a2f98
98DATA ·kcon+0x008(SB)/8, $0x428a2f98428a2f98
99DATA ·kcon+0x010(SB)/8, $0x7137449171374491
100DATA ·kcon+0x018(SB)/8, $0x7137449171374491
101DATA ·kcon+0x020(SB)/8, $0xb5c0fbcfb5c0fbcf
102DATA ·kcon+0x028(SB)/8, $0xb5c0fbcfb5c0fbcf
103DATA ·kcon+0x030(SB)/8, $0xe9b5dba5e9b5dba5
104DATA ·kcon+0x038(SB)/8, $0xe9b5dba5e9b5dba5
105DATA ·kcon+0x040(SB)/8, $0x3956c25b3956c25b
106DATA ·kcon+0x048(SB)/8, $0x3956c25b3956c25b
107DATA ·kcon+0x050(SB)/8, $0x59f111f159f111f1
108DATA ·kcon+0x058(SB)/8, $0x59f111f159f111f1
109DATA ·kcon+0x060(SB)/8, $0x923f82a4923f82a4
110DATA ·kcon+0x068(SB)/8, $0x923f82a4923f82a4
111DATA ·kcon+0x070(SB)/8, $0xab1c5ed5ab1c5ed5
112DATA ·kcon+0x078(SB)/8, $0xab1c5ed5ab1c5ed5
113DATA ·kcon+0x080(SB)/8, $0xd807aa98d807aa98
114DATA ·kcon+0x088(SB)/8, $0xd807aa98d807aa98
115DATA ·kcon+0x090(SB)/8, $0x12835b0112835b01
116DATA ·kcon+0x098(SB)/8, $0x12835b0112835b01
117DATA ·kcon+0x0A0(SB)/8, $0x243185be243185be
118DATA ·kcon+0x0A8(SB)/8, $0x243185be243185be
119DATA ·kcon+0x0B0(SB)/8, $0x550c7dc3550c7dc3
120DATA ·kcon+0x0B8(SB)/8, $0x550c7dc3550c7dc3
121DATA ·kcon+0x0C0(SB)/8, $0x72be5d7472be5d74
122DATA ·kcon+0x0C8(SB)/8, $0x72be5d7472be5d74
123DATA ·kcon+0x0D0(SB)/8, $0x80deb1fe80deb1fe
124DATA ·kcon+0x0D8(SB)/8, $0x80deb1fe80deb1fe
125DATA ·kcon+0x0E0(SB)/8, $0x9bdc06a79bdc06a7
126DATA ·kcon+0x0E8(SB)/8, $0x9bdc06a79bdc06a7
127DATA ·kcon+0x0F0(SB)/8, $0xc19bf174c19bf174
128DATA ·kcon+0x0F8(SB)/8, $0xc19bf174c19bf174
129DATA ·kcon+0x100(SB)/8, $0xe49b69c1e49b69c1
130DATA ·kcon+0x108(SB)/8, $0xe49b69c1e49b69c1
131DATA ·kcon+0x110(SB)/8, $0xefbe4786efbe4786
132DATA ·kcon+0x118(SB)/8, $0xefbe4786efbe4786
133DATA ·kcon+0x120(SB)/8, $0x0fc19dc60fc19dc6
134DATA ·kcon+0x128(SB)/8, $0x0fc19dc60fc19dc6
135DATA ·kcon+0x130(SB)/8, $0x240ca1cc240ca1cc
136DATA ·kcon+0x138(SB)/8, $0x240ca1cc240ca1cc
137DATA ·kcon+0x140(SB)/8, $0x2de92c6f2de92c6f
138DATA ·kcon+0x148(SB)/8, $0x2de92c6f2de92c6f
139DATA ·kcon+0x150(SB)/8, $0x4a7484aa4a7484aa
140DATA ·kcon+0x158(SB)/8, $0x4a7484aa4a7484aa
141DATA ·kcon+0x160(SB)/8, $0x5cb0a9dc5cb0a9dc
142DATA ·kcon+0x168(SB)/8, $0x5cb0a9dc5cb0a9dc
143DATA ·kcon+0x170(SB)/8, $0x76f988da76f988da
144DATA ·kcon+0x178(SB)/8, $0x76f988da76f988da
145DATA ·kcon+0x180(SB)/8, $0x983e5152983e5152
146DATA ·kcon+0x188(SB)/8, $0x983e5152983e5152
147DATA ·kcon+0x190(SB)/8, $0xa831c66da831c66d
148DATA ·kcon+0x198(SB)/8, $0xa831c66da831c66d
149DATA ·kcon+0x1A0(SB)/8, $0xb00327c8b00327c8
150DATA ·kcon+0x1A8(SB)/8, $0xb00327c8b00327c8
151DATA ·kcon+0x1B0(SB)/8, $0xbf597fc7bf597fc7
152DATA ·kcon+0x1B8(SB)/8, $0xbf597fc7bf597fc7
153DATA ·kcon+0x1C0(SB)/8, $0xc6e00bf3c6e00bf3
154DATA ·kcon+0x1C8(SB)/8, $0xc6e00bf3c6e00bf3
155DATA ·kcon+0x1D0(SB)/8, $0xd5a79147d5a79147
156DATA ·kcon+0x1D8(SB)/8, $0xd5a79147d5a79147
157DATA ·kcon+0x1E0(SB)/8, $0x06ca635106ca6351
158DATA ·kcon+0x1E8(SB)/8, $0x06ca635106ca6351
159DATA ·kcon+0x1F0(SB)/8, $0x1429296714292967
160DATA ·kcon+0x1F8(SB)/8, $0x1429296714292967
161DATA ·kcon+0x200(SB)/8, $0x27b70a8527b70a85
162DATA ·kcon+0x208(SB)/8, $0x27b70a8527b70a85
163DATA ·kcon+0x210(SB)/8, $0x2e1b21382e1b2138
164DATA ·kcon+0x218(SB)/8, $0x2e1b21382e1b2138
165DATA ·kcon+0x220(SB)/8, $0x4d2c6dfc4d2c6dfc
166DATA ·kcon+0x228(SB)/8, $0x4d2c6dfc4d2c6dfc
167DATA ·kcon+0x230(SB)/8, $0x53380d1353380d13
168DATA ·kcon+0x238(SB)/8, $0x53380d1353380d13
169DATA ·kcon+0x240(SB)/8, $0x650a7354650a7354
170DATA ·kcon+0x248(SB)/8, $0x650a7354650a7354
171DATA ·kcon+0x250(SB)/8, $0x766a0abb766a0abb
172DATA ·kcon+0x258(SB)/8, $0x766a0abb766a0abb
173DATA ·kcon+0x260(SB)/8, $0x81c2c92e81c2c92e
174DATA ·kcon+0x268(SB)/8, $0x81c2c92e81c2c92e
175DATA ·kcon+0x270(SB)/8, $0x92722c8592722c85
176DATA ·kcon+0x278(SB)/8, $0x92722c8592722c85
177DATA ·kcon+0x280(SB)/8, $0xa2bfe8a1a2bfe8a1
178DATA ·kcon+0x288(SB)/8, $0xa2bfe8a1a2bfe8a1
179DATA ·kcon+0x290(SB)/8, $0xa81a664ba81a664b
180DATA ·kcon+0x298(SB)/8, $0xa81a664ba81a664b
181DATA ·kcon+0x2A0(SB)/8, $0xc24b8b70c24b8b70
182DATA ·kcon+0x2A8(SB)/8, $0xc24b8b70c24b8b70
183DATA ·kcon+0x2B0(SB)/8, $0xc76c51a3c76c51a3
184DATA ·kcon+0x2B8(SB)/8, $0xc76c51a3c76c51a3
185DATA ·kcon+0x2C0(SB)/8, $0xd192e819d192e819
186DATA ·kcon+0x2C8(SB)/8, $0xd192e819d192e819
187DATA ·kcon+0x2D0(SB)/8, $0xd6990624d6990624
188DATA ·kcon+0x2D8(SB)/8, $0xd6990624d6990624
189DATA ·kcon+0x2E0(SB)/8, $0xf40e3585f40e3585
190DATA ·kcon+0x2E8(SB)/8, $0xf40e3585f40e3585
191DATA ·kcon+0x2F0(SB)/8, $0x106aa070106aa070
192DATA ·kcon+0x2F8(SB)/8, $0x106aa070106aa070
193DATA ·kcon+0x300(SB)/8, $0x19a4c11619a4c116
194DATA ·kcon+0x308(SB)/8, $0x19a4c11619a4c116
195DATA ·kcon+0x310(SB)/8, $0x1e376c081e376c08
196DATA ·kcon+0x318(SB)/8, $0x1e376c081e376c08
197DATA ·kcon+0x320(SB)/8, $0x2748774c2748774c
198DATA ·kcon+0x328(SB)/8, $0x2748774c2748774c
199DATA ·kcon+0x330(SB)/8, $0x34b0bcb534b0bcb5
200DATA ·kcon+0x338(SB)/8, $0x34b0bcb534b0bcb5
201DATA ·kcon+0x340(SB)/8, $0x391c0cb3391c0cb3
202DATA ·kcon+0x348(SB)/8, $0x391c0cb3391c0cb3
203DATA ·kcon+0x350(SB)/8, $0x4ed8aa4a4ed8aa4a
204DATA ·kcon+0x358(SB)/8, $0x4ed8aa4a4ed8aa4a
205DATA ·kcon+0x360(SB)/8, $0x5b9cca4f5b9cca4f
206DATA ·kcon+0x368(SB)/8, $0x5b9cca4f5b9cca4f
207DATA ·kcon+0x370(SB)/8, $0x682e6ff3682e6ff3
208DATA ·kcon+0x378(SB)/8, $0x682e6ff3682e6ff3
209DATA ·kcon+0x380(SB)/8, $0x748f82ee748f82ee
210DATA ·kcon+0x388(SB)/8, $0x748f82ee748f82ee
211DATA ·kcon+0x390(SB)/8, $0x78a5636f78a5636f
212DATA ·kcon+0x398(SB)/8, $0x78a5636f78a5636f
213DATA ·kcon+0x3A0(SB)/8, $0x84c8781484c87814
214DATA ·kcon+0x3A8(SB)/8, $0x84c8781484c87814
215DATA ·kcon+0x3B0(SB)/8, $0x8cc702088cc70208
216DATA ·kcon+0x3B8(SB)/8, $0x8cc702088cc70208
217DATA ·kcon+0x3C0(SB)/8, $0x90befffa90befffa
218DATA ·kcon+0x3C8(SB)/8, $0x90befffa90befffa
219DATA ·kcon+0x3D0(SB)/8, $0xa4506ceba4506ceb
220DATA ·kcon+0x3D8(SB)/8, $0xa4506ceba4506ceb
221DATA ·kcon+0x3E0(SB)/8, $0xbef9a3f7bef9a3f7
222DATA ·kcon+0x3E8(SB)/8, $0xbef9a3f7bef9a3f7
223DATA ·kcon+0x3F0(SB)/8, $0xc67178f2c67178f2
224DATA ·kcon+0x3F8(SB)/8, $0xc67178f2c67178f2
225DATA ·kcon+0x400(SB)/8, $0x0000000000000000
226DATA ·kcon+0x408(SB)/8, $0x0000000000000000
227
228#ifdef GOARCH_ppc64le
229DATA ·kcon+0x410(SB)/8, $0x1011121310111213 // permutation control vectors
230DATA ·kcon+0x418(SB)/8, $0x1011121300010203
231DATA ·kcon+0x420(SB)/8, $0x1011121310111213
232DATA ·kcon+0x428(SB)/8, $0x0405060700010203
233DATA ·kcon+0x430(SB)/8, $0x1011121308090a0b
234DATA ·kcon+0x438(SB)/8, $0x0405060700010203
235#else
236DATA ·kcon+0x410(SB)/8, $0x1011121300010203
237DATA ·kcon+0x418(SB)/8, $0x1011121310111213 // permutation control vectors
238DATA ·kcon+0x420(SB)/8, $0x0405060700010203
239DATA ·kcon+0x428(SB)/8, $0x1011121310111213
240DATA ·kcon+0x430(SB)/8, $0x0001020304050607
241DATA ·kcon+0x438(SB)/8, $0x08090a0b10111213
242#endif
243
244GLOBL ·kcon(SB), RODATA, $1088
245
246#define SHA256ROUND0(a, b, c, d, e, f, g, h, xi, idx) \
247 VSEL g, f, e, FUNC; \
248 VSHASIGMAW $15, e, $1, S1; \
249 VADDUWM xi, h, h; \
250 VSHASIGMAW $0, a, $1, S0; \
251 VADDUWM FUNC, h, h; \
252 VXOR b, a, FUNC; \
253 VADDUWM S1, h, h; \
254 VSEL b, c, FUNC, FUNC; \
255 VADDUWM KI, g, g; \
256 VADDUWM h, d, d; \
257 VADDUWM FUNC, S0, S0; \
258 LVX (TBL)(idx), KI; \
259 VADDUWM S0, h, h
260
261#define SHA256ROUND1(a, b, c, d, e, f, g, h, xi, xj, xj_1, xj_9, xj_14, idx) \
262 VSHASIGMAW $0, xj_1, $0, s0; \
263 VSEL g, f, e, FUNC; \
264 VSHASIGMAW $15, e, $1, S1; \
265 VADDUWM xi, h, h; \
266 VSHASIGMAW $0, a, $1, S0; \
267 VSHASIGMAW $15, xj_14, $0, s1; \
268 VADDUWM FUNC, h, h; \
269 VXOR b, a, FUNC; \
270 VADDUWM xj_9, xj, xj; \
271 VADDUWM S1, h, h; \
272 VSEL b, c, FUNC, FUNC; \
273 VADDUWM KI, g, g; \
274 VADDUWM h, d, d; \
275 VADDUWM FUNC, S0, S0; \
276 VADDUWM s0, xj, xj; \
277 LVX (TBL)(idx), KI; \
278 VADDUWM S0, h, h; \
279 VADDUWM s1, xj, xj
280
281#ifdef GOARCH_ppc64le
282#define VPERMLE(va,vb,vc,vt) VPERM va, vb, vc, vt
283#else
284#define VPERMLE(va,vb,vc,vt)
285#endif
286
287// func block(dig *digest, p []byte)
288TEXT ·block(SB),0,$0-32
289 MOVD dig+0(FP), CTX
290 MOVD p_base+8(FP), INP
291 MOVD p_len+16(FP), LEN
292
293 SRD $6, LEN
294 SLD $6, LEN
295 ADD INP, LEN, END
296
297 CMP INP, END
298 BEQ end
299
300 MOVD $·kcon(SB), TBL_STRT
301 MOVD $0x10, R_x010
302
303#ifdef GOARCH_ppc64le
304 MOVWZ $8, TEMP
305 LVSL (TEMP)(R0), LEMASK
306 VSPLTISB $0x0F, KI
307 VXOR KI, LEMASK, LEMASK
308#endif
309
310 LXVW4X (CTX)(R_x000), V0
311 LXVW4X (CTX)(R_x010), V4
312
313 // unpack the input values into vector registers
314 VSLDOI $4, V0, V0, V1
315 VSLDOI $8, V0, V0, V2
316 VSLDOI $12, V0, V0, V3
317 VSLDOI $4, V4, V4, V5
318 VSLDOI $8, V4, V4, V6
319 VSLDOI $12, V4, V4, V7
320
321 MOVD $0x020, R_x020
322 MOVD $0x030, R_x030
323 MOVD $0x040, R_x040
324 MOVD $0x050, R_x050
325 MOVD $0x060, R_x060
326 MOVD $0x070, R_x070
327 MOVD $0x080, R_x080
328 MOVD $0x090, R_x090
329 MOVD $0x0a0, R_x0a0
330 MOVD $0x0b0, R_x0b0
331 MOVD $0x0c0, R_x0c0
332 MOVD $0x0d0, R_x0d0
333 MOVD $0x0e0, R_x0e0
334 MOVD $0x0f0, R_x0f0
335 MOVD $0x100, R_x100
336 MOVD $0x110, R_x110
337
338loop:
339 MOVD TBL_STRT, TBL
340 LVX (TBL)(R_x000), KI
341
342 LXVD2X (INP)(R_x000), V8 // load v8 in advance
343
344 // Offload to VSR24-31 (aka FPR24-31)
345 XXLOR V0, V0, VS24
346 XXLOR V1, V1, VS25
347 XXLOR V2, V2, VS26
348 XXLOR V3, V3, VS27
349 XXLOR V4, V4, VS28
350 XXLOR V5, V5, VS29
351 XXLOR V6, V6, VS30
352 XXLOR V7, V7, VS31
353
354 VADDUWM KI, V7, V7 // h+K[i]
355 LVX (TBL)(R_x010), KI
356
357 VPERMLE(V8, V8, LEMASK, V8)
358 SHA256ROUND0(V0, V1, V2, V3, V4, V5, V6, V7, V8, R_x020)
359 VSLDOI $4, V8, V8, V9
360 SHA256ROUND0(V7, V0, V1, V2, V3, V4, V5, V6, V9, R_x030)
361 VSLDOI $4, V9, V9, V10
362 SHA256ROUND0(V6, V7, V0, V1, V2, V3, V4, V5, V10, R_x040)
363 LXVD2X (INP)(R_x010), V12 // load v12 in advance
364 VSLDOI $4, V10, V10, V11
365 SHA256ROUND0(V5, V6, V7, V0, V1, V2, V3, V4, V11, R_x050)
366 VPERMLE(V12, V12, LEMASK, V12)
367 SHA256ROUND0(V4, V5, V6, V7, V0, V1, V2, V3, V12, R_x060)
368 VSLDOI $4, V12, V12, V13
369 SHA256ROUND0(V3, V4, V5, V6, V7, V0, V1, V2, V13, R_x070)
370 VSLDOI $4, V13, V13, V14
371 SHA256ROUND0(V2, V3, V4, V5, V6, V7, V0, V1, V14, R_x080)
372 LXVD2X (INP)(R_x020), V16 // load v16 in advance
373 VSLDOI $4, V14, V14, V15
374 SHA256ROUND0(V1, V2, V3, V4, V5, V6, V7, V0, V15, R_x090)
375 VPERMLE(V16, V16, LEMASK, V16)
376 SHA256ROUND0(V0, V1, V2, V3, V4, V5, V6, V7, V16, R_x0a0)
377 VSLDOI $4, V16, V16, V17
378 SHA256ROUND0(V7, V0, V1, V2, V3, V4, V5, V6, V17, R_x0b0)
379 VSLDOI $4, V17, V17, V18
380 SHA256ROUND0(V6, V7, V0, V1, V2, V3, V4, V5, V18, R_x0c0)
381 VSLDOI $4, V18, V18, V19
382 LXVD2X (INP)(R_x030), V20 // load v20 in advance
383 SHA256ROUND0(V5, V6, V7, V0, V1, V2, V3, V4, V19, R_x0d0)
384 VPERMLE(V20, V20, LEMASK, V20)
385 SHA256ROUND0(V4, V5, V6, V7, V0, V1, V2, V3, V20, R_x0e0)
386 VSLDOI $4, V20, V20, V21
387 SHA256ROUND0(V3, V4, V5, V6, V7, V0, V1, V2, V21, R_x0f0)
388 VSLDOI $4, V21, V21, V22
389 SHA256ROUND0(V2, V3, V4, V5, V6, V7, V0, V1, V22, R_x100)
390 VSLDOI $4, V22, V22, V23
391 SHA256ROUND1(V1, V2, V3, V4, V5, V6, V7, V0, V23, V8, V9, V17, V22, R_x110)
392
393 MOVD $3, TEMP
394 MOVD TEMP, CTR
395 ADD $0x120, TBL
396 ADD $0x40, INP
397
398L16_xx:
399 SHA256ROUND1(V0, V1, V2, V3, V4, V5, V6, V7, V8, V9, V10, V18, V23, R_x000)
400 SHA256ROUND1(V7, V0, V1, V2, V3, V4, V5, V6, V9, V10, V11, V19, V8, R_x010)
401 SHA256ROUND1(V6, V7, V0, V1, V2, V3, V4, V5, V10, V11, V12, V20, V9, R_x020)
402 SHA256ROUND1(V5, V6, V7, V0, V1, V2, V3, V4, V11, V12, V13, V21, V10, R_x030)
403 SHA256ROUND1(V4, V5, V6, V7, V0, V1, V2, V3, V12, V13, V14, V22, V11, R_x040)
404 SHA256ROUND1(V3, V4, V5, V6, V7, V0, V1, V2, V13, V14, V15, V23, V12, R_x050)
405 SHA256ROUND1(V2, V3, V4, V5, V6, V7, V0, V1, V14, V15, V16, V8, V13, R_x060)
406 SHA256ROUND1(V1, V2, V3, V4, V5, V6, V7, V0, V15, V16, V17, V9, V14, R_x070)
407 SHA256ROUND1(V0, V1, V2, V3, V4, V5, V6, V7, V16, V17, V18, V10, V15, R_x080)
408 SHA256ROUND1(V7, V0, V1, V2, V3, V4, V5, V6, V17, V18, V19, V11, V16, R_x090)
409 SHA256ROUND1(V6, V7, V0, V1, V2, V3, V4, V5, V18, V19, V20, V12, V17, R_x0a0)
410 SHA256ROUND1(V5, V6, V7, V0, V1, V2, V3, V4, V19, V20, V21, V13, V18, R_x0b0)
411 SHA256ROUND1(V4, V5, V6, V7, V0, V1, V2, V3, V20, V21, V22, V14, V19, R_x0c0)
412 SHA256ROUND1(V3, V4, V5, V6, V7, V0, V1, V2, V21, V22, V23, V15, V20, R_x0d0)
413 SHA256ROUND1(V2, V3, V4, V5, V6, V7, V0, V1, V22, V23, V8, V16, V21, R_x0e0)
414 SHA256ROUND1(V1, V2, V3, V4, V5, V6, V7, V0, V23, V8, V9, V17, V22, R_x0f0)
415 ADD $0x100, TBL
416
417 BDNZ L16_xx
418
419 XXLOR VS24, VS24, V10
420
421 XXLOR VS25, VS25, V11
422 VADDUWM V10, V0, V0
423 XXLOR VS26, VS26, V12
424 VADDUWM V11, V1, V1
425 XXLOR VS27, VS27, V13
426 VADDUWM V12, V2, V2
427 XXLOR VS28, VS28, V14
428 VADDUWM V13, V3, V3
429 XXLOR VS29, VS29, V15
430 VADDUWM V14, V4, V4
431 XXLOR VS30, VS30, V16
432 VADDUWM V15, V5, V5
433 XXLOR VS31, VS31, V17
434 VADDUWM V16, V6, V6
435 VADDUWM V17, V7, V7
436
437 CMPU INP, END
438 BLT loop
439
440 LVX (TBL)(R_x000), V8
441 VPERM V0, V1, KI, V0
442 LVX (TBL)(R_x010), V9
443 VPERM V4, V5, KI, V4
444 VPERM V0, V2, V8, V0
445 VPERM V4, V6, V8, V4
446 VPERM V0, V3, V9, V0
447 VPERM V4, V7, V9, V4
448 STXVD2X V0, (CTX+R_x000)
449 STXVD2X V4, (CTX+R_x010)
450
451end:
452 RET
453
View as plain text