diff options
Diffstat (limited to 'vere/ext/gmp/gen/x86_64-macos/mpn/mullo_basecase.s')
| -rw-r--r-- | vere/ext/gmp/gen/x86_64-macos/mpn/mullo_basecase.s | 436 |
1 files changed, 436 insertions, 0 deletions
diff --git a/vere/ext/gmp/gen/x86_64-macos/mpn/mullo_basecase.s b/vere/ext/gmp/gen/x86_64-macos/mpn/mullo_basecase.s new file mode 100644 index 0000000..127dd3f --- /dev/null +++ b/vere/ext/gmp/gen/x86_64-macos/mpn/mullo_basecase.s @@ -0,0 +1,436 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + .text + .align 5, 0x90 + .globl ___gmpn_mullo_basecase + + +___gmpn_mullo_basecase: + + + + mov %rdx, %r8 + mov (%rsi), %rdx + + cmp $4, %rcx + jb Lsmall + + push %rbx + push %rbp + push %r12 + push %r13 + + mov (%r8), %r9 + mov 8(%r8), %rbx + + lea 2(%rcx), %rbp + shr $2, %rbp + neg %rcx + add $2, %rcx + + push %rsi + + test $1, %cl + jnz Lm2x1 + +Lm2x0:.byte 0xc4,66,171,0xf6,233 + xor %r12d, %r12d + test $2, %cl + jz Lm2b2 + +Lm2b0:lea -8(%rdi), %rdi + lea -8(%rsi), %rsi + jmp Lm2e0 + +Lm2b2:lea -24(%rdi), %rdi + lea 8(%rsi), %rsi + jmp Lm2e2 + +Lm2x1:.byte 0xc4,66,155,0xf6,217 + xor %r10d, %r10d + test $2, %cl + jnz Lm2b3 + +Lm2b1:jmp Lm2e1 + +Lm2b3:lea -16(%rdi), %rdi + lea -16(%rsi), %rsi + jmp Lm2e3 + + .align 4, 0x90 +Lm2tp:.byte 0xc4,98,251,0xf6,211 + add %rax, %r12 + mov (%rsi), %rdx + .byte 0xc4,66,251,0xf6,217 + adc $0, %r10 + add %rax, %r12 + adc $0, %r11 + add %r13, %r12 +Lm2e1:mov %r12, (%rdi) + adc $0, %r11 + .byte 0xc4,98,251,0xf6,227 + add %rax, %r10 + mov 8(%rsi), %rdx + adc $0, %r12 + .byte 0xc4,66,251,0xf6,233 + add %rax, %r10 + adc $0, %r13 + add %r11, %r10 +Lm2e0:mov %r10, 8(%rdi) + adc $0, %r13 + .byte 0xc4,98,251,0xf6,211 + add %rax, %r12 + mov 16(%rsi), %rdx + .byte 0xc4,66,251,0xf6,217 + adc $0, %r10 + add %rax, %r12 + adc $0, %r11 + add %r13, %r12 +Lm2e3:mov %r12, 16(%rdi) + adc $0, %r11 + .byte 0xc4,98,251,0xf6,227 + add %rax, %r10 + mov 24(%rsi), %rdx + adc $0, %r12 + .byte 0xc4,66,251,0xf6,233 + add %rax, %r10 + adc $0, %r13 + add %r11, %r10 + lea 32(%rsi), %rsi +Lm2e2:mov %r10, 24(%rdi) + adc $0, %r13 + dec %rbp + lea 32(%rdi), %rdi + jnz Lm2tp + +Lm2ed:.byte 0xc4,98,251,0xf6,211 + add %rax, %r12 + mov (%rsi), %rdx + .byte 0xc4,66,251,0xf6,217 + add %r12, %rax + add %r13, %rax + mov %rax, (%rdi) + + mov (%rsp), %rsi + lea 16(%r8), %r8 + lea 8(%rdi,%rcx,8), %rdi + add $2, %rcx + jge Lcor1 + + push %r14 + push %r15 + +Louter: + mov (%r8), %r9 + mov 8(%r8), %rbx + + lea (%rcx), %rbp + sar $2, %rbp + + mov (%rsi), %rdx + test $1, %cl + jnz Lbx1 + +Lbx0: mov (%rdi), %r15 + mov 8(%rdi), %r14 + .byte 0xc4,66,251,0xf6,233 + add %rax, %r15 + adc $0, %r13 + .byte 0xc4,98,251,0xf6,211 + add %rax, %r14 + adc $0, %r10 + mov 8(%rsi), %rdx + mov %r15, (%rdi) + .byte 0xc4,66,251,0xf6,217 + test $2, %cl + jz Lb2 + +Lb0: lea 8(%rdi), %rdi + lea 8(%rsi), %rsi + jmp Llo0 + +Lb2: mov 16(%rdi), %r15 + lea 24(%rdi), %rdi + lea 24(%rsi), %rsi + jmp Llo2 + +Lbx1: mov (%rdi), %r14 + mov 8(%rdi), %r15 + .byte 0xc4,66,251,0xf6,217 + add %rax, %r14 + .byte 0xc4,98,251,0xf6,227 + adc $0, %r11 + mov %r14, (%rdi) + add %rax, %r15 + adc $0, %r12 + mov 8(%rsi), %rdx + test $2, %cl + jnz Lb3 + +Lb1: lea 16(%rsi), %rsi + lea 16(%rdi), %rdi + jmp Llo1 + +Lb3: mov 16(%rdi), %r14 + lea 32(%rsi), %rsi + .byte 0xc4,66,251,0xf6,233 + inc %rbp + jz Lcj3 + jmp Llo3 + + .align 4, 0x90 +Ltop: .byte 0xc4,66,251,0xf6,233 + add %r10, %r15 + adc $0, %r12 +Llo3: add %rax, %r15 + adc $0, %r13 + .byte 0xc4,98,251,0xf6,211 + add %rax, %r14 + adc $0, %r10 + lea 32(%rdi), %rdi + add %r11, %r15 + mov -16(%rsi), %rdx + mov %r15, -24(%rdi) + adc $0, %r13 + add %r12, %r14 + mov -8(%rdi), %r15 + .byte 0xc4,66,251,0xf6,217 + adc $0, %r10 +Llo2: add %rax, %r14 + .byte 0xc4,98,251,0xf6,227 + adc $0, %r11 + add %r13, %r14 + mov %r14, -16(%rdi) + adc $0, %r11 + add %rax, %r15 + adc $0, %r12 + add %r10, %r15 + mov -8(%rsi), %rdx + adc $0, %r12 +Llo1: .byte 0xc4,66,251,0xf6,233 + add %rax, %r15 + adc $0, %r13 + mov (%rdi), %r14 + .byte 0xc4,98,251,0xf6,211 + add %rax, %r14 + adc $0, %r10 + add %r11, %r15 + mov %r15, -8(%rdi) + adc $0, %r13 + mov (%rsi), %rdx + add %r12, %r14 + .byte 0xc4,66,251,0xf6,217 + adc $0, %r10 +Llo0: add %rax, %r14 + adc $0, %r11 + .byte 0xc4,98,251,0xf6,227 + add %r13, %r14 + mov 8(%rdi), %r15 + mov %r14, (%rdi) + mov 16(%rdi), %r14 + adc $0, %r11 + add %rax, %r15 + adc $0, %r12 + mov 8(%rsi), %rdx + lea 32(%rsi), %rsi + inc %rbp + jnz Ltop + +Lend: .byte 0xc4,66,251,0xf6,233 + add %r10, %r15 + adc $0, %r12 +Lcj3: add %rax, %r15 + adc $0, %r13 + .byte 0xc4,98,251,0xf6,211 + add %rax, %r14 + add %r11, %r15 + mov -16(%rsi), %rdx + mov %r15, 8(%rdi) + adc $0, %r13 + add %r12, %r14 + .byte 0xc4,66,251,0xf6,217 + add %r14, %rax + add %r13, %rax + mov %rax, 16(%rdi) + + mov 16(%rsp), %rsi + lea 16(%r8), %r8 + lea 24(%rdi,%rcx,8), %rdi + add $2, %rcx + jl Louter + + pop %r15 + pop %r14 + + jnz Lcor0 + +Lcor1:mov (%r8), %r9 + mov 8(%r8), %rbx + mov (%rsi), %rdx + .byte 0xc4,194,155,0xf6,233 + add (%rdi), %r12 + adc %rax, %rbp + mov 8(%rsi), %r10 + imul %r9, %r10 + imul %rbx, %rdx + mov %r12, (%rdi) + add %r10, %rdx + add %rbp, %rdx + mov %rdx, 8(%rdi) + pop %rax + pop %r13 + pop %r12 + pop %rbp + pop %rbx + + ret + +Lcor0:mov (%r8), %r11 + imul (%rsi), %r11 + add %rax, %r11 + mov %r11, (%rdi) + pop %rax + pop %r13 + pop %r12 + pop %rbp + pop %rbx + + ret + + .align 4, 0x90 +Lsmall: + cmp $2, %rcx + jae Lgt1 +Ln1: imul (%r8), %rdx + mov %rdx, (%rdi) + + ret +Lgt1: ja Lgt2 +Ln2: mov (%r8), %r9 + .byte 0xc4,194,251,0xf6,209 + mov %rax, (%rdi) + mov 8(%rsi), %rax + imul %r9, %rax + add %rax, %rdx + mov 8(%r8), %r9 + mov (%rsi), %rcx + imul %r9, %rcx + add %rcx, %rdx + mov %rdx, 8(%rdi) + + ret +Lgt2: +Ln3: mov (%r8), %r9 + .byte 0xc4,66,251,0xf6,209 + mov %rax, (%rdi) + mov 8(%rsi), %rdx + .byte 0xc4,194,251,0xf6,209 + imul 16(%rsi), %r9 + add %rax, %r10 + adc %rdx, %r9 + mov 8(%r8), %r11 + mov (%rsi), %rdx + .byte 0xc4,194,251,0xf6,211 + add %rax, %r10 + adc %rdx, %r9 + imul 8(%rsi), %r11 + add %r11, %r9 + mov %r10, 8(%rdi) + mov 16(%r8), %r10 + mov (%rsi), %rax + imul %rax, %r10 + add %r10, %r9 + mov %r9, 16(%rdi) + + ret + |