Lines Matching full:w
169 # Extract w[t-7]
170 MY_VPALIGNR YTMP0, Y_3, Y_2, 8 # YTMP0 = W[-7]
171 # Calculate w[t-16] + w[t-7]
172 vpaddq Y_0, YTMP0, YTMP0 # YTMP0 = W[-7] + W[-16]
173 # Extract w[t-15]
174 MY_VPALIGNR YTMP1, Y_1, Y_0, 8 # YTMP1 = W[-15]
178 # Calculate w[t-15] ror 1
181 vpor YTMP2, YTMP3, YTMP3 # YTMP3 = W[-15] ror 1
182 # Calculate w[t-15] shr 7
183 vpsrlq $7, YTMP1, YTMP4 # YTMP4 = W[-15] >> 7
188 add frame_XFER(%rsp),h # h = k + w + h # --
200 add h, d # d = k + w + h + d # --
213 add y1, h # h = k + w + h + S0 # --
215 add y2, d # d = k + w + h + d + S1 + CH = d + t1 # --
217 add y2, h # h = k + w + h + S0 + S1 + CH = t1 + S0# --
224 # Calculate w[t-15] ror 8
227 vpor YTMP2, YTMP1, YTMP1 # YTMP1 = W[-15] ror 8
229 vpxor YTMP4, YTMP3, YTMP3 # YTMP3 = W[-15] ror 1 ^ W[-15] >> 7
233 # Add three components, w[t-16], w[t-7] and sigma0
234 vpaddq YTMP1, YTMP0, YTMP0 # YTMP0 = W[-16] + W[-7] + s0
235 # Move to appropriate lanes for calculating w[16] and w[17]
236 vperm2f128 $0x0, YTMP0, YTMP0, Y_0 # Y_0 = W[-16] + W[-7] + s0 {BABA}
237 # Move to appropriate lanes for calculating w[18] and w[19]
238 vpand MASK_YMM_LO(%rip), YTMP0, YTMP0 # YTMP0 = W[-16] + W[-7] + s0 {DC00}
240 # Calculate w[16] and w[17] in both 128 bit lanes
242 # Calculate sigma1 for w[16] and w[17] on both 128 bit lanes
243 vperm2f128 $0x11, Y_3, Y_3, YTMP2 # YTMP2 = W[-2] {BABA}
244 vpsrlq $6, YTMP2, YTMP4 # YTMP4 = W[-2] >> 6 {BABA}
250 add 1*8+frame_XFER(%rsp), h # h = k + w + h # --
264 add h, d # d = k + w + h + d # --
278 add y1, h # h = k + w + h + S0 # --
280 add y2, d # d = k + w + h + d + S1 + CH = d + t1 # --
281 add y2, h # h = k + w + h + S0 + S1 + CH = t1 + S0# --
289 vpsrlq $19, YTMP2, YTMP3 # YTMP3 = W[-2] >> 19 {BABA}
290 vpsllq $(64-19), YTMP2, YTMP1 # YTMP1 = W[-2] << 19 {BABA}
291 vpor YTMP1, YTMP3, YTMP3 # YTMP3 = W[-2] ror 19 {BABA}
292 vpxor YTMP3, YTMP4, YTMP4 # YTMP4 = W[-2] ror 19 ^ W[-2] >> 6 {BABA}
293 vpsrlq $61, YTMP2, YTMP3 # YTMP3 = W[-2] >> 61 {BABA}
294 vpsllq $(64-61), YTMP2, YTMP1 # YTMP1 = W[-2] << 61 {BABA}
295 vpor YTMP1, YTMP3, YTMP3 # YTMP3 = W[-2] ror 61 {BABA}
296 vpxor YTMP3, YTMP4, YTMP4 # YTMP4 = s1 = (W[-2] ror 19) ^
297 # (W[-2] ror 61) ^ (W[-2] >> 6) {BABA}
299 # Add sigma1 to the other compunents to get w[16] and w[17]
300 vpaddq YTMP4, Y_0, Y_0 # Y_0 = {W[1], W[0], W[1], W[0]}
302 # Calculate sigma1 for w[18] and w[19] for upper 128 bit lane
303 vpsrlq $6, Y_0, YTMP4 # YTMP4 = W[-2] >> 6 {DC--}
307 add 2*8+frame_XFER(%rsp), h # h = k + w + h # --
319 add h, d # d = k + w + h + d # --
335 add y1, h # h = k + w + h + S0 # --
336 add y2, d # d = k + w + h + d + S1 + CH = d + t1 # --
337 add y2, h # h = k + w + h + S0 + S1 + CH = t1 + S0# --
345 vpsrlq $19, Y_0, YTMP3 # YTMP3 = W[-2] >> 19 {DC--}
346 vpsllq $(64-19), Y_0, YTMP1 # YTMP1 = W[-2] << 19 {DC--}
347 vpor YTMP1, YTMP3, YTMP3 # YTMP3 = W[-2] ror 19 {DC--}
348 vpxor YTMP3, YTMP4, YTMP4 # YTMP4 = W[-2] ror 19 ^ W[-2] >> 6 {DC--}
349 vpsrlq $61, Y_0, YTMP3 # YTMP3 = W[-2] >> 61 {DC--}
350 vpsllq $(64-61), Y_0, YTMP1 # YTMP1 = W[-2] << 61 {DC--}
351 vpor YTMP1, YTMP3, YTMP3 # YTMP3 = W[-2] ror 61 {DC--}
352 vpxor YTMP3, YTMP4, YTMP4 # YTMP4 = s1 = (W[-2] ror 19) ^
353 # (W[-2] ror 61) ^ (W[-2] >> 6) {DC--}
355 # Add the sigma0 + w[t-7] + w[t-16] for w[18] and w[19]
356 # to newly calculated sigma1 to get w[18] and w[19]
357 vpaddq YTMP4, YTMP0, YTMP2 # YTMP2 = {W[3], W[2], --, --}
359 # Form w[19, w[18], w17], w[16]
360 vpblendd $0xF0, YTMP2, Y_0, Y_0 # Y_0 = {W[3], W[2], W[1], W[0]}
365 add 3*8+frame_XFER(%rsp), h # h = k + w + h # --
377 add h, d # d = k + w + h + d # --
387 add y2, d # d = k + w + h + d + S1 + CH = d + t1 # --
396 add y1, h # h = k + w + h + S0 # --
397 add y2, h # h = k + w + h + S0 + S1 + CH = t1 + S0# --
426 add frame_XFER(%rsp), h # h = k + w + h # --
435 add h, d # d = k + w + h + d # --
437 add y1, h # h = k + w + h + S0 # --
439 add y2, d # d = k + w + h + d + S1 + CH = d + t1 # --
445 add y2, old_h # h = k + w + h + S0 + S1 + CH = t1 + S0# --
464 add 8*1+frame_XFER(%rsp), h # h = k + w + h # --
473 add h, d # d = k + w + h + d # --
475 add y1, h # h = k + w + h + S0 # --
477 add y2, d # d = k + w + h + d + S1 + CH = d + t1 # --
483 add y2, old_h # h = k + w + h + S0 + S1 + CH = t1 + S0# --
502 add 8*2+frame_XFER(%rsp), h # h = k + w + h # --
511 add h, d # d = k + w + h + d # --
513 add y1, h # h = k + w + h + S0 # --
515 add y2, d # d = k + w + h + d + S1 + CH = d + t1 # --
521 add y2, old_h # h = k + w + h + S0 + S1 + CH = t1 + S0# --
540 add 8*3+frame_XFER(%rsp), h # h = k + w + h # --
550 add h, d # d = k + w + h + d # --
552 add y1, h # h = k + w + h + S0 # --
554 add y2, d # d = k + w + h + d + S1 + CH = d + t1 # --
556 add y2, h # h = k + w + h + S0 + S1 + CH = t1 + S0# --