Home
last modified time | relevance | path

Searched refs:acc3 (Results 1 – 25 of 73) sorted by relevance

123

/third_party/openssl/crypto/ec/asm/
Decp_nistz256-armv8.pl50 $acc0,$acc1,$acc2,$acc3,$acc4,$acc5) =
230 ldp $acc2,$acc3,[$ap,#16]
252 ldp $acc2,$acc3,[$ap,#16]
273 ldp $acc2,$acc3,[$ap,#16]
279 mov $t3,$acc3
298 ldp $acc2,$acc3,[$ap,#16]
304 mov $t3,$acc3
308 mov $a3,$acc3
335 ldp $acc2,$acc3,[$ap,#16]
359 mov $acc3,xzr
[all …]
Decp_nistz256-ppc64.pl45 my ($rp,$ap,$bp,$bi,$acc0,$acc1,$acc2,$acc3,$poly1,$poly3,
217 ld $acc3,24($ap)
253 ld $acc3,24($ap)
288 ld $acc3,24($ap)
293 mr $t3,$acc3
328 ld $acc3,24($ap)
336 mr $t3,$acc3
337 std $acc3,88($sp)
380 ld $acc3,24($ap)
416 li $acc3,0
[all …]
Decp_nistz256-x86_64.pl492 my ($acc0,$acc1,$acc2,$acc3,$acc4,$acc5,$acc6,$acc7)=map("%r$_",(8..15));
556 mov %rdx, $acc3
558 add %rax, $acc3
585 sub %rax, $acc3
589 add $acc0, $acc3
610 add $t1, $acc3
612 add %rax, $acc3
636 sub $acc1, $acc3
644 adc %rdx, $acc3
667 add $t1, $acc3
[all …]
Dx25519-x86_64.pl485 my ($acc0,$acc1,$acc2,$acc3,$acc4,$acc5,$acc6,$acc7) = map("%r$_",(8..15));
539 mulx $acc7,$acc3,$acc4 # a[0]*b[3]
541 adcx %rax,$acc3
550 adcx %rbx,$acc3
552 adox %rax,$acc3
562 adox %rbx,$acc3
564 adcx %rax,$acc3
576 adox %rax,$acc3
629 mulx %rsi,$acc3,$acc4 # a[0]*a[3]
631 adcx %rbx,$acc3
[all …]
Decp_nistz256-sparcv9.pl1603 my ($acc0,$acc1,$acc2,$acc3,$acc4,$acc5)=map("%o$_",(0..5));
1614 addxccc $acc3,$acc3,$acc3
1632 addxccc $t3,$acc3,$acc3
1640 addxccc $acc3,$poly3,$t3
1648 movrz $acc4,$t3,$acc3
1651 stx $acc3,[$rp+24]
1657 ! $acc0-$acc3 to 32-bit values and pick b[4] in 32-bit pieces. But
1660 ! "collect" result back to 64-bit $acc0-$acc3.
1686 srlx $acc3,32,$acc5
1689 subccc $acc3,$t2,$acc3
[all …]
Dx25519-ppc64.pl50 $acc0,$acc1,$acc2,$acc3,$acc4,$acc5,$acc6,$acc7) =
87 mulld $acc3,$a3,$bi # a[3]*b[0]
90 for(my @acc=($acc0,$acc1,$acc2,$acc3,$acc4,$acc5,$acc6,$acc7),
132 adde $acc3,$acc3,$t3
138 adde $acc3,$acc3,$t2
146 addze $acc3,$acc3
153 std $acc3,24($rp)
214 mulld $acc3,$a3,$a0 # a[3]*a[0]
220 adde $acc3,$acc3,$t2
232 addc $acc3,$acc3,$t0 # accumulate low parts of multiplication
[all …]
/third_party/openssl/crypto/bn/asm/
Darmv8-mont.pl663 my ($acc0,$acc1,$acc2,$acc3,$acc4,$acc5,$acc6,$acc7)=map("x$_",(19..26));
715 mov $acc3,xzr
763 adcs $acc3,$acc3,$t2
777 adcs $acc3,$acc3,$t0
790 adds $acc3,$acc3,$t2
802 stp $acc2,$acc3,[$tp],#8*2 // t[2..3]
849 adc $acc3,xzr,xzr // t[11]
856 adc $acc3,$acc3,$t3
863 adcs $acc3,$acc3,$t2
868 adcs $acc3,$acc3,$t0
[all …]
Dppc-mont.pl360 $acc0,$acc1,$acc2,$acc3,$acc4,
436 li $acc3,0
466 adde $acc3,$acc3,$t3
476 adde $acc3,$acc3,$t2
496 adde $acc2,$acc3,$t3
498 adde $acc3,$acc4,$carry
503 adde $acc3,$acc3,$t3
536 adde $acc3,$acc3,$t3
544 adde $acc3,$acc3,$t2
554 adde $acc3,$acc3,$t3
[all …]
/third_party/cmsis/CMSIS/DSP/Source/MatrixFunctions/
Darm_mat_cmplx_mult_q31.c78 q63_t acc0, acc1, acc2, acc3; in arm_mat_cmplx_mult_q31_2x2_mve() local
97 acc3 = vmlaldavxq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_2x2_mve()
102 pOut[1 * CMPLX_DIM * MATRIX_DIM2 + 1] = (q31_t) asrl(acc3, 31); in arm_mat_cmplx_mult_q31_2x2_mve()
115 acc3 = vmlaldavxq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_2x2_mve()
122 pOut[1 * CMPLX_DIM * MATRIX_DIM2 + 1] = (q31_t) asrl(acc3, 31); in arm_mat_cmplx_mult_q31_2x2_mve()
141 q63_t acc0, acc1, acc2, acc3; in arm_mat_cmplx_mult_q31_3x3_mve() local
171 acc3 = vmlaldavxq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_3x3_mve()
179 acc3 = vmlaldavaxq_s32(acc3, vecA, vecB1); in arm_mat_cmplx_mult_q31_3x3_mve()
184 pOut[1 * CMPLX_DIM * MATRIX_DIM3 + 1] = (q31_t) asrl(acc3, 31); in arm_mat_cmplx_mult_q31_3x3_mve()
212 acc3 = vmlaldavxq_s32(vecA, vecB); in arm_mat_cmplx_mult_q31_3x3_mve()
[all …]
Darm_mat_cmplx_mult_f32.c305 f32x4_t acc0, acc1, acc2, acc3; in arm_mat_cmplx_mult_f32_4x4_mve() local
335 acc3 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f32_4x4_mve()
336 acc3 = vcmlaq_rot90(acc3, vecA, vecB); in arm_mat_cmplx_mult_f32_4x4_mve()
353 acc3 = vcmlaq(acc3, vecA, vecB); in arm_mat_cmplx_mult_f32_4x4_mve()
354 acc3 = vcmlaq_rot90(acc3, vecA, vecB); in arm_mat_cmplx_mult_f32_4x4_mve()
362 pOut[3 * CMPLX_DIM * MATRIX_DIM4 + 0] = acc3[0] + acc3[2]; in arm_mat_cmplx_mult_f32_4x4_mve()
363 pOut[3 * CMPLX_DIM * MATRIX_DIM4 + 1] = acc3[1] + acc3[3]; in arm_mat_cmplx_mult_f32_4x4_mve()
386 acc3 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f32_4x4_mve()
387 acc3 = vcmlaq_rot90(acc3, vecA, vecB); in arm_mat_cmplx_mult_f32_4x4_mve()
404 acc3 = vcmlaq(acc3, vecA, vecB); in arm_mat_cmplx_mult_f32_4x4_mve()
[all …]
Darm_mat_mult_q31.c225 q63_t acc0, acc1, acc2, acc3; in arm_mat_mult_q31_4x4_mve() local
241 acc3 = vrmlaldavhq(vecA, vecB); in arm_mat_mult_q31_4x4_mve()
246 acc3 = asrl(acc3, 23); in arm_mat_mult_q31_4x4_mve()
251 pOut[3 * MATRIX_DIM4] = (q31_t) acc3; in arm_mat_mult_q31_4x4_mve()
266 acc3 = vrmlaldavhq(vecA, vecB); in arm_mat_mult_q31_4x4_mve()
271 acc3 = asrl(acc3, 23); in arm_mat_mult_q31_4x4_mve()
276 pOut[3 * MATRIX_DIM4] = (q31_t) acc3; in arm_mat_mult_q31_4x4_mve()
292 acc3 = vrmlaldavhq(vecA, vecB); in arm_mat_mult_q31_4x4_mve()
297 acc3 = asrl(acc3, 23); in arm_mat_mult_q31_4x4_mve()
302 pOut[3 * MATRIX_DIM4] = (q31_t) acc3; in arm_mat_mult_q31_4x4_mve()
[all …]
Darm_mat_cmplx_mult_f16.c248 f16x8_t acc0, acc1, acc2, acc3; in arm_mat_cmplx_mult_f16_4x4_mve() local
275 acc3 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve()
276 acc3 = vcmlaq_rot90(acc3, vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve()
282 mve_cmplx_sum_intra_vec_f16(acc3, &pOut[3 * CMPLX_DIM * MATRIX_DIM]); in arm_mat_cmplx_mult_f16_4x4_mve()
304 acc3 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve()
305 acc3 = vcmlaq_rot90(acc3, vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve()
311 mve_cmplx_sum_intra_vec_f16(acc3, &pOut[3 * CMPLX_DIM * MATRIX_DIM]); in arm_mat_cmplx_mult_f16_4x4_mve()
333 acc3 = vcmulq(vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve()
334 acc3 = vcmlaq_rot90(acc3, vecA, vecB); in arm_mat_cmplx_mult_f16_4x4_mve()
340 mve_cmplx_sum_intra_vec_f16(acc3, &pOut[3 * CMPLX_DIM * MATRIX_DIM]); in arm_mat_cmplx_mult_f16_4x4_mve()
[all …]
/third_party/openssl/crypto/aes/asm/
Daes-sparcv9.pl49 $acc3="%o2";
227 sll $s3,3,$acc3
231 and $acc3,2040,$acc3
235 ldx [$tbl+$acc3],$acc3
284 srlx $acc3,24,$acc3
288 xor $acc3,$t0,$t0
318 sll $t3,3,$acc3
323 and $acc3,2040,$acc3
327 ldx [$tbl+$acc3],$acc3
367 srlx $acc3,24,$acc3
[all …]
Daes-parisc.pl67 ($acc0, $acc1, $acc2, $acc3, $acc4, $acc5, $acc6, $acc7,
217 _srm $s3,0,$acc3
222 ldwx,s $acc3($tbl),$acc3
254 _ror $acc3,24,$acc3
258 xor $acc3,$t0,$t0
285 _srm $t3,0,$acc3
290 ldwx,s $acc3($tbl),$acc3
317 _ror $acc3,24,$acc3
324 xor $acc3,$s0,$s0
367 _ror $acc3,24,$acc3
[all …]
/third_party/cmsis/CMSIS/DSP/Source/FilteringFunctions/
Darm_conv_fast_q15.c73 q31_t sum, acc0, acc1, acc2, acc3; /* Accumulators */ in arm_conv_fast_q15() local
282 acc3 = 0; in arm_conv_fast_q15()
317 acc3 = __SMLADX(x3, c0, acc3); in arm_conv_fast_q15()
339 acc3 = __SMLADX(x1, c0, acc3); in arm_conv_fast_q15()
369 acc3 = __SMLADX(x3, c0, acc3); in arm_conv_fast_q15()
388 acc3 = __SMLADX(x2, c0, acc3); in arm_conv_fast_q15()
406 acc3 = __SMLADX(x2, c0, acc3); in arm_conv_fast_q15()
424 acc3 = __SMLADX(x3, c0, acc3); in arm_conv_fast_q15()
430 write_q15x2_ia (&pOut, __PKHBT((acc2 >> 15), (acc3 >> 15), 16)); in arm_conv_fast_q15()
433 write_q15x2_ia (&pOut, __PKHBT((acc3 >> 15), (acc2 >> 15), 16)); in arm_conv_fast_q15()
Darm_correlate_fast_q15.c73 q31_t sum, acc0, acc1, acc2, acc3; /* Accumulators */ in arm_correlate_fast_q15() local
268 acc3 = 0; in arm_correlate_fast_q15()
303 acc3 = __SMLAD(x3, c0, acc3); in arm_correlate_fast_q15()
325 acc3 = __SMLAD(x1, c0, acc3); in arm_correlate_fast_q15()
355 acc3 = __SMLADX(x3, c0, acc3); in arm_correlate_fast_q15()
374 acc3 = __SMLAD(x2, c0, acc3); in arm_correlate_fast_q15()
392 acc3 = __SMLAD(x2, c0, acc3); in arm_correlate_fast_q15()
410 acc3 = __SMLADX(x3, c0, acc3); in arm_correlate_fast_q15()
424 *pOut = (q15_t) (acc3 >> 15); in arm_correlate_fast_q15()
Darm_fir_decimate_fast_q31.c81 q63_t acc1, acc2, acc3; in arm_fir_decimate_fast_q31() local
109 acc3 = 0; in arm_fir_decimate_fast_q31()
141 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_fir_decimate_fast_q31()
156 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_fir_decimate_fast_q31()
171 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_fir_decimate_fast_q31()
186 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_fir_decimate_fast_q31()
210 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_fir_decimate_fast_q31()
224 *pDst++ = (q31_t) (acc3 << 1); in arm_fir_decimate_fast_q31()
Darm_conv_partial_fast_q15.c68 q31_t sum, acc0, acc1, acc2, acc3; /* Accumulator */ in arm_conv_partial_fast_q15() local
304 acc3 = 0; in arm_conv_partial_fast_q15()
341 acc3 = __SMLADX(x3, c0, acc3); in arm_conv_partial_fast_q15()
363 acc3 = __SMLADX(x1, c0, acc3); in arm_conv_partial_fast_q15()
392 acc3 = __SMLADX(x3, c0, acc3); in arm_conv_partial_fast_q15()
411 acc3 = __SMLADX(x2, c0, acc3); in arm_conv_partial_fast_q15()
429 acc3 = __SMLADX(x2, c0, acc3); in arm_conv_partial_fast_q15()
446 acc3 = __SMLADX(x3, c0, acc3); in arm_conv_partial_fast_q15()
452 write_q15x2_ia (&pOut, __PKHBT(acc2 >> 15, acc3 >> 15, 16)); in arm_conv_partial_fast_q15()
455 write_q15x2_ia (&pOut, __PKHBT(acc3 >> 15, acc2 >> 15, 16)); in arm_conv_partial_fast_q15()
Darm_fir_fast_q15.c76 q31_t acc1, acc2, acc3; /* Accumulators */ in arm_fir_fast_q15() local
108 acc3 = 0; in arm_fir_fast_q15()
158 acc3 = __SMLADX(x1, c0, acc3); in arm_fir_fast_q15()
183 acc3 = __SMLADX(x1, c0, acc3); in arm_fir_fast_q15()
221 acc3 = __SMLADX(x1, c0, acc3); in arm_fir_fast_q15()
228 write_q15x2_ia (&pDst, __PKHBT(__SSAT((acc2 >> 15), 16), __SSAT((acc3 >> 15), 16), 16)); in arm_fir_fast_q15()
231 write_q15x2_ia (&pDst, __PKHBT(__SSAT((acc3 >> 15), 16), __SSAT((acc2 >> 15), 16), 16)); in arm_fir_fast_q15()
Darm_fir_q31.c235 q63_t acc0=0, acc1=0, acc2=0, acc3=0; in arm_fir_q31_1_4_mve() local
263 acc3 = vrmlaldavhq(vecIn0, vecCoeffs); in arm_fir_q31_1_4_mve()
268 acc3 = asrl(acc3, 23); in arm_fir_q31_1_4_mve()
273 *pOutput++ = (q31_t) acc3; in arm_fir_q31_1_4_mve()
465 q63_t acc0, acc1, acc2, acc3; in arm_fir_q31_29_32_mve() local
561 acc3 = vrmlaldavhq(vecIn0, vecCoeffs0); in arm_fir_q31_29_32_mve()
563 acc3 = vrmlaldavhaq(acc3, vecIn0, vecCoeffs1); in arm_fir_q31_29_32_mve()
565 acc3 = vrmlaldavhaq(acc3, vecIn0, vecCoeffs2); in arm_fir_q31_29_32_mve()
567 acc3 = vrmlaldavhaq(acc3, vecIn0, vecCoeffs3); in arm_fir_q31_29_32_mve()
569 acc3 = vrmlaldavhaq(acc3, vecIn0, vecCoeffs4); in arm_fir_q31_29_32_mve()
[all …]
Darm_conv_fast_q31.c76 q31_t sum, acc0, acc1, acc2, acc3; /* Accumulators */ in arm_conv_fast_q31() local
245 acc3 = 0; in arm_conv_fast_q31()
272 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_conv_fast_q31()
288 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x0 * c0)) >> 32); in arm_conv_fast_q31()
304 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x1 * c0)) >> 32); in arm_conv_fast_q31()
320 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x2 * c0)) >> 32); in arm_conv_fast_q31()
344 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_conv_fast_q31()
359 *pOut++ = (q31_t) (acc3 << 1); in arm_conv_fast_q31()
Darm_conv_partial_q15.c74 q63_t sum, acc0, acc1, acc2, acc3; /* Accumulator */ in arm_conv_partial_q15() local
310 acc3 = 0; in arm_conv_partial_q15()
347 acc3 = __SMLALDX(x3, c0, acc3); in arm_conv_partial_q15()
369 acc3 = __SMLALDX(x1, c0, acc3); in arm_conv_partial_q15()
398 acc3 = __SMLALDX(x3, c0, acc3); in arm_conv_partial_q15()
417 acc3 = __SMLALDX(x2, c0, acc3); in arm_conv_partial_q15()
435 acc3 = __SMLALDX(x2, c0, acc3); in arm_conv_partial_q15()
452 acc3 = __SMLALDX(x3, c0, acc3); in arm_conv_partial_q15()
458 write_q15x2_ia (&pOut, __PKHBT(__SSAT((acc2 >> 15), 16), __SSAT((acc3 >> 15), 16), 16)); in arm_conv_partial_q15()
461 write_q15x2_ia (&pOut, __PKHBT(__SSAT((acc3 >> 15), 16), __SSAT((acc2 >> 15), 16), 16)); in arm_conv_partial_q15()
Darm_conv_q15.c146 int64_t acc3 = 0LL; in arm_conv_q15() local
155 MVE_INTR_CONV_QUAD_INC_X_FIXED_SIZE_Q15(acc0, acc1, acc2, acc3, pX, pY, count); in arm_conv_q15()
159 *pDst++ = (q15_t) acc3; in arm_conv_q15()
236 q63_t sum, acc0, acc1, acc2, acc3; /* Accumulators */ in arm_conv_q15() local
443 acc3 = 0; in arm_conv_q15()
479 acc3 = __SMLALDX(x3, c0, acc3); in arm_conv_q15()
502 acc3 = __SMLALDX(x1, c0, acc3); in arm_conv_q15()
531 acc3 = __SMLALDX(x3, c0, acc3); in arm_conv_q15()
550 acc3 = __SMLALDX(x2, c0, acc3); in arm_conv_q15()
568 acc3 = __SMLALDX(x2, c0, acc3); in arm_conv_q15()
[all …]
Darm_correlate_fast_q31.c78 q31_t sum, acc0, acc1, acc2, acc3; /* Accumulators */ in arm_correlate_fast_q31() local
271 acc3 = 0; in arm_correlate_fast_q31()
298 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_correlate_fast_q31()
314 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x0 * c0)) >> 32); in arm_correlate_fast_q31()
330 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x1 * c0)) >> 32); in arm_correlate_fast_q31()
346 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x2 * c0)) >> 32); in arm_correlate_fast_q31()
370 acc3 = (q31_t) ((((q63_t) acc3 << 32) + ((q63_t) x3 * c0)) >> 32); in arm_correlate_fast_q31()
392 *pOut = (q31_t) (acc3 << 1); in arm_correlate_fast_q31()
/third_party/mbedtls/library/
Dpoly1305.c100 uint32_t acc0, acc1, acc2, acc3, acc4; in poly1305_process() local
118 acc3 = ctx->acc[3]; in poly1305_process()
134 d3 += (uint64_t) acc3 + ( d2 >> 32U ); in poly1305_process()
138 acc3 = (uint32_t) d3; in poly1305_process()
145 mul64( acc3, rs1 ); in poly1305_process()
149 mul64( acc3, rs2 ) + in poly1305_process()
154 mul64( acc3, rs3 ) + in poly1305_process()
159 mul64( acc3, r0 ) + in poly1305_process()
170 acc3 = (uint32_t) d3; in poly1305_process()
180 d0 = (uint64_t) acc3 + ( d0 >> 32U ); in poly1305_process()
[all …]

123