123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446 |
- # File: dclxvi-20130329/fp2e_mulxi.s
- # Author: Ruben Niederhagen, Peter Schwabe
- # Public Domain
- # qhasm: enter fp2e_mulxi_qhasm
- .text
- .p2align 5
- .globl _fp2e_mulxi_qhasm
- .globl fp2e_mulxi_qhasm
- _fp2e_mulxi_qhasm:
- fp2e_mulxi_qhasm:
- mov %rsp,%r11
- and $31,%r11
- add $0,%r11
- sub %r11,%rsp
- # qhasm: int64 0rop
- # qhasm: int64 0op
- # qhasm: input 0rop
- # qhasm: input 0op
- # qhasm: int6464 0r0
- # qhasm: int6464 0r1
- # qhasm: int6464 0r2
- # qhasm: int6464 0r3
- # qhasm: int6464 0r4
- # qhasm: int6464 0r5
- # qhasm: int6464 0r6
- # qhasm: int6464 0r7
- # qhasm: int6464 0r8
- # qhasm: int6464 0r9
- # qhasm: int6464 0r10
- # qhasm: int6464 0r11
- # qhasm: int6464 0t0
- # qhasm: int6464 0t1
- # qhasm: int6464 0t2
- # qhasm: int6464 0t3
- # qhasm: 0r0 = *(int128 *)(0op + 0)
- # asm 1: movdqa 0(<0op=int64#2),>0r0=int6464#1
- # asm 2: movdqa 0(<0op=%rsi),>0r0=%xmm0
- movdqa 0(%rsi),%xmm0
- # qhasm: 0r1 = *(int128 *)(0op + 16)
- # asm 1: movdqa 16(<0op=int64#2),>0r1=int6464#2
- # asm 2: movdqa 16(<0op=%rsi),>0r1=%xmm1
- movdqa 16(%rsi),%xmm1
- # qhasm: 0r2 = *(int128 *)(0op + 32)
- # asm 1: movdqa 32(<0op=int64#2),>0r2=int6464#3
- # asm 2: movdqa 32(<0op=%rsi),>0r2=%xmm2
- movdqa 32(%rsi),%xmm2
- # qhasm: 0r3 = *(int128 *)(0op + 48)
- # asm 1: movdqa 48(<0op=int64#2),>0r3=int6464#4
- # asm 2: movdqa 48(<0op=%rsi),>0r3=%xmm3
- movdqa 48(%rsi),%xmm3
- # qhasm: 0r4 = *(int128 *)(0op + 64)
- # asm 1: movdqa 64(<0op=int64#2),>0r4=int6464#5
- # asm 2: movdqa 64(<0op=%rsi),>0r4=%xmm4
- movdqa 64(%rsi),%xmm4
- # qhasm: 0r5 = *(int128 *)(0op + 80)
- # asm 1: movdqa 80(<0op=int64#2),>0r5=int6464#6
- # asm 2: movdqa 80(<0op=%rsi),>0r5=%xmm5
- movdqa 80(%rsi),%xmm5
- # qhasm: 0r6 = *(int128 *)(0op + 96)
- # asm 1: movdqa 96(<0op=int64#2),>0r6=int6464#7
- # asm 2: movdqa 96(<0op=%rsi),>0r6=%xmm6
- movdqa 96(%rsi),%xmm6
- # qhasm: 0r7 = *(int128 *)(0op + 112)
- # asm 1: movdqa 112(<0op=int64#2),>0r7=int6464#8
- # asm 2: movdqa 112(<0op=%rsi),>0r7=%xmm7
- movdqa 112(%rsi),%xmm7
- # qhasm: 0r8 = *(int128 *)(0op + 128)
- # asm 1: movdqa 128(<0op=int64#2),>0r8=int6464#9
- # asm 2: movdqa 128(<0op=%rsi),>0r8=%xmm8
- movdqa 128(%rsi),%xmm8
- # qhasm: 0r9 = *(int128 *)(0op + 144)
- # asm 1: movdqa 144(<0op=int64#2),>0r9=int6464#10
- # asm 2: movdqa 144(<0op=%rsi),>0r9=%xmm9
- movdqa 144(%rsi),%xmm9
- # qhasm: 0r10 = *(int128 *)(0op + 160)
- # asm 1: movdqa 160(<0op=int64#2),>0r10=int6464#11
- # asm 2: movdqa 160(<0op=%rsi),>0r10=%xmm10
- movdqa 160(%rsi),%xmm10
- # qhasm: 0r11 = *(int128 *)(0op + 176)
- # asm 1: movdqa 176(<0op=int64#2),>0r11=int6464#12
- # asm 2: movdqa 176(<0op=%rsi),>0r11=%xmm11
- movdqa 176(%rsi),%xmm11
- # qhasm: int6464 1t0
- # qhasm: int6464 1t1
- # qhasm: int6464 1t2
- # qhasm: int6464 1t3
- # qhasm: int6464 0t4
- # qhasm: int6464 0t5
- # qhasm: int6464 0t6
- # qhasm: int6464 0t7
- # qhasm: int6464 0t8
- # qhasm: int6464 0t9
- # qhasm: int6464 0t10
- # qhasm: int6464 0t11
- # qhasm: 1t0 = 0r0
- # asm 1: movdqa <0r0=int6464#1,>1t0=int6464#13
- # asm 2: movdqa <0r0=%xmm0,>1t0=%xmm12
- movdqa %xmm0,%xmm12
- # qhasm: float6464 0r0 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r0=int6464#1
- # asm 2: mulpd THREE_MINUSONE,<0r0=%xmm0
- mulpd THREE_MINUSONE,%xmm0
- # qhasm: float6464 1t0 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<1t0=int6464#13
- # asm 2: mulpd ONE_THREE,<1t0=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r0[0] += 0r0[1];0r0[1] = 1t0[0] + 1t0[1]
- # asm 1: haddpd <1t0=int6464#13,<0r0=int6464#1
- # asm 2: haddpd <1t0=%xmm12,<0r0=%xmm0
- haddpd %xmm12,%xmm0
- # qhasm: 1t1 = 0r1
- # asm 1: movdqa <0r1=int6464#2,>1t1=int6464#13
- # asm 2: movdqa <0r1=%xmm1,>1t1=%xmm12
- movdqa %xmm1,%xmm12
- # qhasm: float6464 0r1 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r1=int6464#2
- # asm 2: mulpd THREE_MINUSONE,<0r1=%xmm1
- mulpd THREE_MINUSONE,%xmm1
- # qhasm: float6464 1t1 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<1t1=int6464#13
- # asm 2: mulpd ONE_THREE,<1t1=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r1[0] += 0r1[1];0r1[1] = 1t1[0] + 1t1[1]
- # asm 1: haddpd <1t1=int6464#13,<0r1=int6464#2
- # asm 2: haddpd <1t1=%xmm12,<0r1=%xmm1
- haddpd %xmm12,%xmm1
- # qhasm: 1t2 = 0r2
- # asm 1: movdqa <0r2=int6464#3,>1t2=int6464#13
- # asm 2: movdqa <0r2=%xmm2,>1t2=%xmm12
- movdqa %xmm2,%xmm12
- # qhasm: float6464 0r2 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r2=int6464#3
- # asm 2: mulpd THREE_MINUSONE,<0r2=%xmm2
- mulpd THREE_MINUSONE,%xmm2
- # qhasm: float6464 1t2 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<1t2=int6464#13
- # asm 2: mulpd ONE_THREE,<1t2=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r2[0] += 0r2[1];0r2[1] = 1t2[0] + 1t2[1]
- # asm 1: haddpd <1t2=int6464#13,<0r2=int6464#3
- # asm 2: haddpd <1t2=%xmm12,<0r2=%xmm2
- haddpd %xmm12,%xmm2
- # qhasm: 1t3 = 0r3
- # asm 1: movdqa <0r3=int6464#4,>1t3=int6464#13
- # asm 2: movdqa <0r3=%xmm3,>1t3=%xmm12
- movdqa %xmm3,%xmm12
- # qhasm: float6464 0r3 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r3=int6464#4
- # asm 2: mulpd THREE_MINUSONE,<0r3=%xmm3
- mulpd THREE_MINUSONE,%xmm3
- # qhasm: float6464 1t3 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<1t3=int6464#13
- # asm 2: mulpd ONE_THREE,<1t3=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r3[0] += 0r3[1];0r3[1] = 1t3[0] + 1t3[1]
- # asm 1: haddpd <1t3=int6464#13,<0r3=int6464#4
- # asm 2: haddpd <1t3=%xmm12,<0r3=%xmm3
- haddpd %xmm12,%xmm3
- # qhasm: 0t4 = 0r4
- # asm 1: movdqa <0r4=int6464#5,>0t4=int6464#13
- # asm 2: movdqa <0r4=%xmm4,>0t4=%xmm12
- movdqa %xmm4,%xmm12
- # qhasm: float6464 0r4 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r4=int6464#5
- # asm 2: mulpd THREE_MINUSONE,<0r4=%xmm4
- mulpd THREE_MINUSONE,%xmm4
- # qhasm: float6464 0t4 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<0t4=int6464#13
- # asm 2: mulpd ONE_THREE,<0t4=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r4[0] += 0r4[1];0r4[1] = 0t4[0] + 0t4[1]
- # asm 1: haddpd <0t4=int6464#13,<0r4=int6464#5
- # asm 2: haddpd <0t4=%xmm12,<0r4=%xmm4
- haddpd %xmm12,%xmm4
- # qhasm: 0t5 = 0r5
- # asm 1: movdqa <0r5=int6464#6,>0t5=int6464#13
- # asm 2: movdqa <0r5=%xmm5,>0t5=%xmm12
- movdqa %xmm5,%xmm12
- # qhasm: float6464 0r5 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r5=int6464#6
- # asm 2: mulpd THREE_MINUSONE,<0r5=%xmm5
- mulpd THREE_MINUSONE,%xmm5
- # qhasm: float6464 0t5 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<0t5=int6464#13
- # asm 2: mulpd ONE_THREE,<0t5=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r5[0] += 0r5[1];0r5[1] = 0t5[0] + 0t5[1]
- # asm 1: haddpd <0t5=int6464#13,<0r5=int6464#6
- # asm 2: haddpd <0t5=%xmm12,<0r5=%xmm5
- haddpd %xmm12,%xmm5
- # qhasm: 0t6 = 0r6
- # asm 1: movdqa <0r6=int6464#7,>0t6=int6464#13
- # asm 2: movdqa <0r6=%xmm6,>0t6=%xmm12
- movdqa %xmm6,%xmm12
- # qhasm: float6464 0r6 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r6=int6464#7
- # asm 2: mulpd THREE_MINUSONE,<0r6=%xmm6
- mulpd THREE_MINUSONE,%xmm6
- # qhasm: float6464 0t6 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<0t6=int6464#13
- # asm 2: mulpd ONE_THREE,<0t6=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r6[0] += 0r6[1];0r6[1] = 0t6[0] + 0t6[1]
- # asm 1: haddpd <0t6=int6464#13,<0r6=int6464#7
- # asm 2: haddpd <0t6=%xmm12,<0r6=%xmm6
- haddpd %xmm12,%xmm6
- # qhasm: 0t7 = 0r7
- # asm 1: movdqa <0r7=int6464#8,>0t7=int6464#13
- # asm 2: movdqa <0r7=%xmm7,>0t7=%xmm12
- movdqa %xmm7,%xmm12
- # qhasm: float6464 0r7 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r7=int6464#8
- # asm 2: mulpd THREE_MINUSONE,<0r7=%xmm7
- mulpd THREE_MINUSONE,%xmm7
- # qhasm: float6464 0t7 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<0t7=int6464#13
- # asm 2: mulpd ONE_THREE,<0t7=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r7[0] += 0r7[1];0r7[1] = 0t7[0] + 0t7[1]
- # asm 1: haddpd <0t7=int6464#13,<0r7=int6464#8
- # asm 2: haddpd <0t7=%xmm12,<0r7=%xmm7
- haddpd %xmm12,%xmm7
- # qhasm: 0t8 = 0r8
- # asm 1: movdqa <0r8=int6464#9,>0t8=int6464#13
- # asm 2: movdqa <0r8=%xmm8,>0t8=%xmm12
- movdqa %xmm8,%xmm12
- # qhasm: float6464 0r8 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r8=int6464#9
- # asm 2: mulpd THREE_MINUSONE,<0r8=%xmm8
- mulpd THREE_MINUSONE,%xmm8
- # qhasm: float6464 0t8 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<0t8=int6464#13
- # asm 2: mulpd ONE_THREE,<0t8=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r8[0] += 0r8[1];0r8[1] = 0t8[0] + 0t8[1]
- # asm 1: haddpd <0t8=int6464#13,<0r8=int6464#9
- # asm 2: haddpd <0t8=%xmm12,<0r8=%xmm8
- haddpd %xmm12,%xmm8
- # qhasm: 0t9 = 0r9
- # asm 1: movdqa <0r9=int6464#10,>0t9=int6464#13
- # asm 2: movdqa <0r9=%xmm9,>0t9=%xmm12
- movdqa %xmm9,%xmm12
- # qhasm: float6464 0r9 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r9=int6464#10
- # asm 2: mulpd THREE_MINUSONE,<0r9=%xmm9
- mulpd THREE_MINUSONE,%xmm9
- # qhasm: float6464 0t9 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<0t9=int6464#13
- # asm 2: mulpd ONE_THREE,<0t9=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r9[0] += 0r9[1];0r9[1] = 0t9[0] + 0t9[1]
- # asm 1: haddpd <0t9=int6464#13,<0r9=int6464#10
- # asm 2: haddpd <0t9=%xmm12,<0r9=%xmm9
- haddpd %xmm12,%xmm9
- # qhasm: 0t10 = 0r10
- # asm 1: movdqa <0r10=int6464#11,>0t10=int6464#13
- # asm 2: movdqa <0r10=%xmm10,>0t10=%xmm12
- movdqa %xmm10,%xmm12
- # qhasm: float6464 0r10 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r10=int6464#11
- # asm 2: mulpd THREE_MINUSONE,<0r10=%xmm10
- mulpd THREE_MINUSONE,%xmm10
- # qhasm: float6464 0t10 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<0t10=int6464#13
- # asm 2: mulpd ONE_THREE,<0t10=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r10[0] += 0r10[1];0r10[1] = 0t10[0] + 0t10[1]
- # asm 1: haddpd <0t10=int6464#13,<0r10=int6464#11
- # asm 2: haddpd <0t10=%xmm12,<0r10=%xmm10
- haddpd %xmm12,%xmm10
- # qhasm: 0t11 = 0r11
- # asm 1: movdqa <0r11=int6464#12,>0t11=int6464#13
- # asm 2: movdqa <0r11=%xmm11,>0t11=%xmm12
- movdqa %xmm11,%xmm12
- # qhasm: float6464 0r11 *= THREE_MINUSONE
- # asm 1: mulpd THREE_MINUSONE,<0r11=int6464#12
- # asm 2: mulpd THREE_MINUSONE,<0r11=%xmm11
- mulpd THREE_MINUSONE,%xmm11
- # qhasm: float6464 0t11 *= ONE_THREE
- # asm 1: mulpd ONE_THREE,<0t11=int6464#13
- # asm 2: mulpd ONE_THREE,<0t11=%xmm12
- mulpd ONE_THREE,%xmm12
- # qhasm: float6464 0r11[0] += 0r11[1];0r11[1] = 0t11[0] + 0t11[1]
- # asm 1: haddpd <0t11=int6464#13,<0r11=int6464#12
- # asm 2: haddpd <0t11=%xmm12,<0r11=%xmm11
- haddpd %xmm12,%xmm11
- # qhasm: *(int128 *)(0rop + 0) = 0r0
- # asm 1: movdqa <0r0=int6464#1,0(<0rop=int64#1)
- # asm 2: movdqa <0r0=%xmm0,0(<0rop=%rdi)
- movdqa %xmm0,0(%rdi)
- # qhasm: *(int128 *)(0rop + 16) = 0r1
- # asm 1: movdqa <0r1=int6464#2,16(<0rop=int64#1)
- # asm 2: movdqa <0r1=%xmm1,16(<0rop=%rdi)
- movdqa %xmm1,16(%rdi)
- # qhasm: *(int128 *)(0rop + 32) = 0r2
- # asm 1: movdqa <0r2=int6464#3,32(<0rop=int64#1)
- # asm 2: movdqa <0r2=%xmm2,32(<0rop=%rdi)
- movdqa %xmm2,32(%rdi)
- # qhasm: *(int128 *)(0rop + 48) = 0r3
- # asm 1: movdqa <0r3=int6464#4,48(<0rop=int64#1)
- # asm 2: movdqa <0r3=%xmm3,48(<0rop=%rdi)
- movdqa %xmm3,48(%rdi)
- # qhasm: *(int128 *)(0rop + 64) = 0r4
- # asm 1: movdqa <0r4=int6464#5,64(<0rop=int64#1)
- # asm 2: movdqa <0r4=%xmm4,64(<0rop=%rdi)
- movdqa %xmm4,64(%rdi)
- # qhasm: *(int128 *)(0rop + 80) = 0r5
- # asm 1: movdqa <0r5=int6464#6,80(<0rop=int64#1)
- # asm 2: movdqa <0r5=%xmm5,80(<0rop=%rdi)
- movdqa %xmm5,80(%rdi)
- # qhasm: *(int128 *)(0rop + 96) = 0r6
- # asm 1: movdqa <0r6=int6464#7,96(<0rop=int64#1)
- # asm 2: movdqa <0r6=%xmm6,96(<0rop=%rdi)
- movdqa %xmm6,96(%rdi)
- # qhasm: *(int128 *)(0rop + 112) = 0r7
- # asm 1: movdqa <0r7=int6464#8,112(<0rop=int64#1)
- # asm 2: movdqa <0r7=%xmm7,112(<0rop=%rdi)
- movdqa %xmm7,112(%rdi)
- # qhasm: *(int128 *)(0rop + 128) = 0r8
- # asm 1: movdqa <0r8=int6464#9,128(<0rop=int64#1)
- # asm 2: movdqa <0r8=%xmm8,128(<0rop=%rdi)
- movdqa %xmm8,128(%rdi)
- # qhasm: *(int128 *)(0rop + 144) = 0r9
- # asm 1: movdqa <0r9=int6464#10,144(<0rop=int64#1)
- # asm 2: movdqa <0r9=%xmm9,144(<0rop=%rdi)
- movdqa %xmm9,144(%rdi)
- # qhasm: *(int128 *)(0rop + 160) = 0r10
- # asm 1: movdqa <0r10=int6464#11,160(<0rop=int64#1)
- # asm 2: movdqa <0r10=%xmm10,160(<0rop=%rdi)
- movdqa %xmm10,160(%rdi)
- # qhasm: *(int128 *)(0rop + 176) = 0r11
- # asm 1: movdqa <0r11=int6464#12,176(<0rop=int64#1)
- # asm 2: movdqa <0r11=%xmm11,176(<0rop=%rdi)
- movdqa %xmm11,176(%rdi)
- # qhasm: leave
- add %r11,%rsp
- mov %rdi,%rax
- mov %rsi,%rdx
- ret
|