Lines Matching refs:rsi
344 # 512-bit mul source: [rsi+8*n]
347 # Clobbers all regs except: rcx, rsi, rdi
353 &MULSTEP_512([map("%r$_",(8..15))], "(+8*0)(%rcx)", "%rsi", "%rbp", "%rbx");
357 &MULSTEP_512([map("%r$_",(9..15,8))], "(+8*1)(%rcx)", "%rsi", "%rbp", "%rbx");
426 mov (+$pData_offset+$STACK_DEPTH)(%rsp), %rsi # pM1 (Bsrc) 512 bits, 8 qwords
427 add \$$M1, %rsi
432 &MULADD_256x512("%rdi", "%rcx", "%rsi", "%rbp", "%rbx", \@X); # rotates @X 4 times
461 # rsi -> M1
468 add \$`$M2-$M1`, %rsi # rsi -> M2
474 call MULADD_128x512 # args in rcx, rdi / rbp, rsi, r15-r8
488 add \$`$K1-$M2`, %rsi # rsi -> pK1 ; 128 bits, 2 qwords
490 # MUL_128x128t128 rdi, rcx, rsi ; Q = X2 * K1 (bottom half)
491 # B1:B0 = rsi[1:0] = K1[1:0]
494 mov (%rsi), %r8 # B0
495 mov (+8*1)(%rsi), %rbx # B1
513 sub \$`$K1-$M`, %rsi
518 call MULADD_128x512 # args in rcx, rdi / rbp, rsi, r15-r8
523 mov (+8*0)(%rsi), %rax
524 mov (+8*1)(%rsi), %rbx
525 mov (+8*2)(%rsi), %rdi
526 mov (+8*3)(%rsi), %rdx
540 mov (+$pData_offset+$STACK_DEPTH)(%rsp), %rcx # rsi -> Data (and points to T)
543 # rsi will be used to generate a mask after the addition
544 xor %rsi, %rsi
555 # if there is a carry: rsi = 0xFFFFFFFFFFFFFFFF
556 # if carry is clear: rsi = 0x0000000000000000
557 sbb \$0, %rsi
560 and %rsi, %rax
561 and %rsi, %rbx
562 and %rsi, %rdi
563 and %rsi, %rdx
584 # use the rsi mask as before
586 and %rsi, %rax
587 and %rsi, %rbx
588 and %rsi, %rdi
589 and %rsi, %rdx
602 mov (+$red_result_addr_offset+$STACK_DEPTH)(%rsp), %rsi
603 mov $X[0], (+8*0)(%rsi)
604 mov $X[1], (+8*1)(%rsi)
605 mov $X[2], (+8*2)(%rsi)
606 mov $X[3], (+8*3)(%rsi)
607 mov $X[4], (+8*4)(%rsi)
608 mov $X[5], (+8*5)(%rsi)
609 mov $X[6], (+8*6)(%rsi)
610 mov $X[7], (+8*7)(%rsi)
673 # src2: Address of source 2: rsi
683 # For multiply: dst = rcx, src1 = rdi, src2 = rsi
686 &MUL_512x512("%rsp+$tmp16_offset+8", "%rdi", "%rsi", [map("%r$_",(10..15,8..9))], "%rbp", "%rbx");
1084 &SQR_512("%rsp+$tmp16_offset+8", "%rcx", [map("%r$_",(10..15,8..9))], "%rbx", "%rbp", "%rsi", "%rdi");
1126 mov %rsi, (+$pG_offset)(%rsp)
1133 movdqu (+16*0)(%rsi), %xmm0
1134 movdqu (+16*1)(%rsi), %xmm1
1135 movdqu (+16*2)(%rsi), %xmm2
1136 movdqu (+16*3)(%rsi), %xmm3
1170 mov %rcx, %rsi # pTmp
1184 # rsi -> pTmp
1185 mov %rsi, (+$red_result_addr_offset)(%rsp)
1186 mov (+8*0)(%rsi), %r10
1187 mov (+8*1)(%rsi), %r11
1188 mov (+8*2)(%rsi), %r12
1189 mov (+8*3)(%rsi), %r13
1190 mov (+8*4)(%rsi), %r14
1191 mov (+8*5)(%rsi), %r15
1192 mov (+8*6)(%rsi), %r8
1193 mov (+8*7)(%rsi), %r9
1197 lea (+$tmp_offset)(%rsp), %rsi
1201 mov %rsi, %rcx # rcx = rsi = addr of tmp
1227 lea (+$garray_offset)(%rsp,%rax,2), %rsi
1231 &unswizzle("%rdx", "%rsi", "%rbp", "%rbx", "%rax");
1269 lea (+$garray_offset)(%rsp,%rdx,2), %rsi
1274 &unswizzle("%rdx", "%rsi", "%rbp", "%rbx", "%rax");
1281 mov (+$pResult_offset)(%rsp), %rsi
1338 mov (+8*0)(%rax), %rsi
1342 cmovnc %r8, %rsi
1346 mov %rsi, (+8*0)(%rax)
1351 mov (+8*4)(%rax), %rsi
1355 cmovnc %r12, %rsi
1359 mov %rsi, (+8*4)(%rax)
1364 mov (+$rsp_offset)(%rsp), %rsi
1365 mov 0(%rsi),%r15
1366 mov 8(%rsi),%r14
1367 mov 16(%rsi),%r13
1368 mov 24(%rsi),%r12
1369 mov 32(%rsi),%rbx
1370 mov 40(%rsi),%rbp
1371 lea 48(%rsi),%rsp
1390 push %rsi
1432 mov 16(%rax),%rsi
1434 mov %rsi,168($context) # restore context->Rsi
1438 mov $context,%rsi # context
1442 mov $disp,%rsi
1444 mov 8(%rsi),%rdx # arg2, disp->ImageBase
1445 mov 0(%rsi),%r8 # arg3, disp->ControlPc
1446 mov 16(%rsi),%r9 # arg4, disp->FunctionEntry
1447 mov 40(%rsi),%r10 # disp->ContextRecord
1448 lea 56(%rsi),%r11 # &disp->HandlerData
1449 lea 24(%rsi),%r12 # &disp->EstablisherFrame
1466 pop %rsi
