From 29f1d3668772edaa81ff60e5dd69ee275e1fd75d Mon Sep 17 00:00:00 2001 From: Sunil K Pandey Date: Mon, 7 Mar 2022 10:47:08 -0800 Subject: [PATCH] x86_64: Fix svml_s_acoshf16_core_avx512.S code formatting This commit contains following formatting changes 1. Instructions proceeded by a tab. 2. Instruction less than 8 characters in length have a tab between it and the first operand. 3. Instruction greater than 7 characters in length have a space between it and the first operand. 4. Tabs after `#define`d names and their value. 5. 8 space at the beginning of line replaced by tab. 6. Indent comments with code. 7. Remove redundent .text section. 8. 1 space between line content and line comment. 9. Space after all commas. Reviewed-by: Noah Goldstein --- .../fpu/multiarch/svml_s_acoshf16_core_avx512.S | 743 ++++++++++----------- 1 file changed, 371 insertions(+), 372 deletions(-) diff --git a/sysdeps/x86_64/fpu/multiarch/svml_s_acoshf16_core_avx512.S b/sysdeps/x86_64/fpu/multiarch/svml_s_acoshf16_core_avx512.S index 0ae0e89..3f44e75 100644 --- a/sysdeps/x86_64/fpu/multiarch/svml_s_acoshf16_core_avx512.S +++ b/sysdeps/x86_64/fpu/multiarch/svml_s_acoshf16_core_avx512.S @@ -36,414 +36,413 @@ /* Offsets for data table __svml_sacosh_data_internal_avx512 */ -#define Log_tbl_H 0 -#define Log_tbl_L 128 -#define One 256 -#define SmallThreshold 320 -#define Threshold 384 -#define LargeThreshold 448 -#define ca1 512 -#define c2s 576 -#define c1s 640 -#define AddB5 704 -#define RcpBitMask 768 -#define OneEighth 832 -#define Four 896 -#define poly_coeff3 960 -#define poly_coeff2 1024 -#define poly_coeff1 1088 -#define L2H 1152 -#define L2L 1216 +#define Log_tbl_H 0 +#define Log_tbl_L 128 +#define One 256 +#define SmallThreshold 320 +#define Threshold 384 +#define LargeThreshold 448 +#define ca1 512 +#define c2s 576 +#define c1s 640 +#define AddB5 704 +#define RcpBitMask 768 +#define OneEighth 832 +#define Four 896 +#define poly_coeff3 960 +#define poly_coeff2 1024 +#define poly_coeff1 1088 +#define L2H 1152 +#define L2L 1216 #include - .text - .section .text.exex512,"ax",@progbits + .section .text.exex512, "ax", @progbits ENTRY(_ZGVeN16v_acoshf_skx) - pushq %rbp - cfi_def_cfa_offset(16) - movq %rsp, %rbp - cfi_def_cfa(6, 16) - cfi_offset(6, -16) - andq $-64, %rsp - subq $192, %rsp - vmovups One+__svml_sacosh_data_internal_avx512(%rip), %zmm1 + pushq %rbp + cfi_def_cfa_offset(16) + movq %rsp, %rbp + cfi_def_cfa(6, 16) + cfi_offset(6, -16) + andq $-64, %rsp + subq $192, %rsp + vmovups One+__svml_sacosh_data_internal_avx512(%rip), %zmm1 -/* - * sqrt(1+x^2) ~ Sh + Sl + Sh*Eh*poly_s - * poly_s = c1+c2*Eh - */ - vmovups c2s+__svml_sacosh_data_internal_avx512(%rip), %zmm13 - vmovups c1s+__svml_sacosh_data_internal_avx512(%rip), %zmm15 - -/* polynomial computation for small inputs */ - vmovups ca1+__svml_sacosh_data_internal_avx512(%rip), %zmm9 - -/* very large inputs ? */ - vmovups Threshold+__svml_sacosh_data_internal_avx512(%rip), %zmm10 - -/* out of range inputs? */ - vmovups LargeThreshold+__svml_sacosh_data_internal_avx512(%rip), %zmm11 - -/* not a very small input ? */ - vmovups SmallThreshold+__svml_sacosh_data_internal_avx512(%rip), %zmm6 - vmovaps %zmm0, %zmm8 - -/* x^2 - 1 */ - vmovaps %zmm1, %zmm7 - vfmsub231ps {rn-sae}, %zmm8, %zmm8, %zmm7 - vcmpps $21, {sae}, %zmm10, %zmm8, %k2 - vcmpps $22, {sae}, %zmm11, %zmm8, %k0 - vcmpps $18, {sae}, %zmm1, %zmm8, %k1 - vrsqrt14ps %zmm7, %zmm12 - vcmpps $21, {sae}, %zmm6, %zmm7, %k3 - vmulps {rn-sae}, %zmm9, %zmm7, %zmm4 - -/* Sh ~sqrt(-1+x^2) */ - vmulps {rn-sae}, %zmm12, %zmm7, %zmm5 - -/* Sh+x */ - vaddps {rn-sae}, %zmm8, %zmm5, %zmm9 - -/* (Yh*R0)_low */ - vmovaps %zmm7, %zmm0 - korw %k0, %k1, %k0 - -/* rel. error term: Eh=1-Sh*R0 */ - vmovaps %zmm1, %zmm14 - vfmsub213ps {rn-sae}, %zmm5, %zmm12, %zmm0 - vfnmadd231ps {rn-sae}, %zmm5, %zmm12, %zmm14 - -/* rel. error term: Eh=(1-Sh*R0)-Sl*R0 */ - vfnmadd231ps {rn-sae}, %zmm0, %zmm12, %zmm14 - -/* Sh*Eh */ - vmulps {rn-sae}, %zmm14, %zmm5, %zmm3 - vfmadd231ps {rn-sae}, %zmm14, %zmm13, %zmm15 - -/* Sl + Sh*Eh*poly_s */ - vfmadd213ps {rn-sae}, %zmm0, %zmm15, %zmm3 - -/* Shh */ - vsubps {rn-sae}, %zmm8, %zmm9, %zmm15 + /* + * sqrt(1+x^2) ~ Sh + Sl + Sh*Eh*poly_s + * poly_s = c1+c2*Eh + */ + vmovups c2s+__svml_sacosh_data_internal_avx512(%rip), %zmm13 + vmovups c1s+__svml_sacosh_data_internal_avx512(%rip), %zmm15 + + /* polynomial computation for small inputs */ + vmovups ca1+__svml_sacosh_data_internal_avx512(%rip), %zmm9 -/* polynomial computation for small inputs */ - vaddps {rn-sae}, %zmm3, %zmm5, %zmm0 + /* very large inputs ? */ + vmovups Threshold+__svml_sacosh_data_internal_avx512(%rip), %zmm10 + + /* out of range inputs? */ + vmovups LargeThreshold+__svml_sacosh_data_internal_avx512(%rip), %zmm11 + + /* not a very small input ? */ + vmovups SmallThreshold+__svml_sacosh_data_internal_avx512(%rip), %zmm6 + vmovaps %zmm0, %zmm8 + + /* x^2 - 1 */ + vmovaps %zmm1, %zmm7 + vfmsub231ps {rn-sae}, %zmm8, %zmm8, %zmm7 + vcmpps $21, {sae}, %zmm10, %zmm8, %k2 + vcmpps $22, {sae}, %zmm11, %zmm8, %k0 + vcmpps $18, {sae}, %zmm1, %zmm8, %k1 + vrsqrt14ps %zmm7, %zmm12 + vcmpps $21, {sae}, %zmm6, %zmm7, %k3 + vmulps {rn-sae}, %zmm9, %zmm7, %zmm4 + + /* Sh ~sqrt(-1+x^2) */ + vmulps {rn-sae}, %zmm12, %zmm7, %zmm5 + + /* Sh+x */ + vaddps {rn-sae}, %zmm8, %zmm5, %zmm9 + + /* (Yh*R0)_low */ + vmovaps %zmm7, %zmm0 + korw %k0, %k1, %k0 + + /* rel. error term: Eh=1-Sh*R0 */ + vmovaps %zmm1, %zmm14 + vfmsub213ps {rn-sae}, %zmm5, %zmm12, %zmm0 + vfnmadd231ps {rn-sae}, %zmm5, %zmm12, %zmm14 + + /* rel. error term: Eh=(1-Sh*R0)-Sl*R0 */ + vfnmadd231ps {rn-sae}, %zmm0, %zmm12, %zmm14 + + /* Sh*Eh */ + vmulps {rn-sae}, %zmm14, %zmm5, %zmm3 + vfmadd231ps {rn-sae}, %zmm14, %zmm13, %zmm15 + + /* Sl + Sh*Eh*poly_s */ + vfmadd213ps {rn-sae}, %zmm0, %zmm15, %zmm3 + + /* Shh */ + vsubps {rn-sae}, %zmm8, %zmm9, %zmm15 + + /* polynomial computation for small inputs */ + vaddps {rn-sae}, %zmm3, %zmm5, %zmm0 + + /* Xin0+Sl+Sh*Eh*poly_s ~ x+sqrt(1+x^2) */ + vaddps {rn-sae}, %zmm3, %zmm9, %zmm2 + + /* Shl */ + vsubps {rn-sae}, %zmm15, %zmm5, %zmm10 + vfmadd231ps {rn-sae}, %zmm0, %zmm4, %zmm0 + + /* fixup for very large inputs */ + vmovups OneEighth+__svml_sacosh_data_internal_avx512(%rip), %zmm4 + + /* Sl_high */ + vsubps {rn-sae}, %zmm9, %zmm2, %zmm5 + + /* polynomial */ + vmovups poly_coeff3+__svml_sacosh_data_internal_avx512(%rip), %zmm9 + vmulps {rn-sae}, %zmm4, %zmm8, %zmm2{%k2} + + /* -K*L2L + Tl */ + vmovups L2L+__svml_sacosh_data_internal_avx512(%rip), %zmm4 + + /* Sl_l */ + vsubps {rn-sae}, %zmm5, %zmm3, %zmm3 + vrcp14ps %zmm2, %zmm11 + vmovups Log_tbl_L+__svml_sacosh_data_internal_avx512(%rip), %zmm5 + + /* Xin_low */ + vaddps {rn-sae}, %zmm10, %zmm3, %zmm13 + + /* round reciprocal to 1+4b mantissas */ + vpaddd AddB5+__svml_sacosh_data_internal_avx512(%rip), %zmm11, %zmm12 + vmovups poly_coeff1+__svml_sacosh_data_internal_avx512(%rip), %zmm10 + vandps RcpBitMask+__svml_sacosh_data_internal_avx512(%rip), %zmm12, %zmm14 -/* Xin0+Sl+Sh*Eh*poly_s ~ x+sqrt(1+x^2) */ - vaddps {rn-sae}, %zmm3, %zmm9, %zmm2 + /* fixup for very large inputs */ + vxorps %zmm13, %zmm13, %zmm13{%k2} -/* Shl */ - vsubps {rn-sae}, %zmm15, %zmm5, %zmm10 - vfmadd231ps {rn-sae}, %zmm0, %zmm4, %zmm0 + /* reduced argument for log(): (Rcp*Xin-1)+Rcp*Xin_low */ + vfmsub231ps {rn-sae}, %zmm14, %zmm2, %zmm1 -/* fixup for very large inputs */ - vmovups OneEighth+__svml_sacosh_data_internal_avx512(%rip), %zmm4 + /* exponents */ + vgetexpps {sae}, %zmm14, %zmm12 + vmovups Four+__svml_sacosh_data_internal_avx512(%rip), %zmm2 -/* Sl_high */ - vsubps {rn-sae}, %zmm9, %zmm2, %zmm5 + /* Prepare table index */ + vpsrld $18, %zmm14, %zmm3 + vfmadd231ps {rn-sae}, %zmm14, %zmm13, %zmm1 + vmovups poly_coeff2+__svml_sacosh_data_internal_avx512(%rip), %zmm13 -/* polynomial */ - vmovups poly_coeff3+__svml_sacosh_data_internal_avx512(%rip), %zmm9 - vmulps {rn-sae}, %zmm4, %zmm8, %zmm2{%k2} + /* Table lookups */ + vmovups __svml_sacosh_data_internal_avx512(%rip), %zmm14 + vsubps {rn-sae}, %zmm2, %zmm12, %zmm12{%k2} + vpermt2ps Log_tbl_L+64+__svml_sacosh_data_internal_avx512(%rip), %zmm3, %zmm5 + vpermt2ps Log_tbl_H+64+__svml_sacosh_data_internal_avx512(%rip), %zmm3, %zmm14 -/* -K*L2L + Tl */ - vmovups L2L+__svml_sacosh_data_internal_avx512(%rip), %zmm4 + /* R^2 */ + vmulps {rn-sae}, %zmm1, %zmm1, %zmm11 -/* Sl_l */ - vsubps {rn-sae}, %zmm5, %zmm3, %zmm3 - vrcp14ps %zmm2, %zmm11 - vmovups Log_tbl_L+__svml_sacosh_data_internal_avx512(%rip), %zmm5 + /* -K*L2H + Th */ + vmovups L2H+__svml_sacosh_data_internal_avx512(%rip), %zmm2 + vfmadd231ps {rn-sae}, %zmm1, %zmm9, %zmm13 + vfnmadd231ps {rn-sae}, %zmm12, %zmm2, %zmm14 + vfnmadd213ps {rn-sae}, %zmm5, %zmm4, %zmm12 + vfmadd213ps {rn-sae}, %zmm10, %zmm1, %zmm13 -/* Xin_low */ - vaddps {rn-sae}, %zmm10, %zmm3, %zmm13 + /* Tl + R^2*Poly */ + vfmadd213ps {rn-sae}, %zmm12, %zmm11, %zmm13 -/* round reciprocal to 1+4b mantissas */ - vpaddd AddB5+__svml_sacosh_data_internal_avx512(%rip), %zmm11, %zmm12 - vmovups poly_coeff1+__svml_sacosh_data_internal_avx512(%rip), %zmm10 - vandps RcpBitMask+__svml_sacosh_data_internal_avx512(%rip), %zmm12, %zmm14 + /* R+Tl + R^2*Poly */ + vaddps {rn-sae}, %zmm1, %zmm13, %zmm1 + vaddps {rn-sae}, %zmm1, %zmm14, %zmm0{%k3} -/* fixup for very large inputs */ - vxorps %zmm13, %zmm13, %zmm13{%k2} + /* Go to special inputs processing branch */ + jne L(SPECIAL_VALUES_BRANCH) + # LOE rbx r12 r13 r14 r15 k0 zmm0 zmm8 -/* reduced argument for log(): (Rcp*Xin-1)+Rcp*Xin_low */ - vfmsub231ps {rn-sae}, %zmm14, %zmm2, %zmm1 - -/* exponents */ - vgetexpps {sae}, %zmm14, %zmm12 - vmovups Four+__svml_sacosh_data_internal_avx512(%rip), %zmm2 - -/* Prepare table index */ - vpsrld $18, %zmm14, %zmm3 - vfmadd231ps {rn-sae}, %zmm14, %zmm13, %zmm1 - vmovups poly_coeff2+__svml_sacosh_data_internal_avx512(%rip), %zmm13 - -/* Table lookups */ - vmovups __svml_sacosh_data_internal_avx512(%rip), %zmm14 - vsubps {rn-sae}, %zmm2, %zmm12, %zmm12{%k2} - vpermt2ps Log_tbl_L+64+__svml_sacosh_data_internal_avx512(%rip), %zmm3, %zmm5 - vpermt2ps Log_tbl_H+64+__svml_sacosh_data_internal_avx512(%rip), %zmm3, %zmm14 - -/* R^2 */ - vmulps {rn-sae}, %zmm1, %zmm1, %zmm11 - -/* -K*L2H + Th */ - vmovups L2H+__svml_sacosh_data_internal_avx512(%rip), %zmm2 - vfmadd231ps {rn-sae}, %zmm1, %zmm9, %zmm13 - vfnmadd231ps {rn-sae}, %zmm12, %zmm2, %zmm14 - vfnmadd213ps {rn-sae}, %zmm5, %zmm4, %zmm12 - vfmadd213ps {rn-sae}, %zmm10, %zmm1, %zmm13 - -/* Tl + R^2*Poly */ - vfmadd213ps {rn-sae}, %zmm12, %zmm11, %zmm13 - -/* R+Tl + R^2*Poly */ - vaddps {rn-sae}, %zmm1, %zmm13, %zmm1 - vaddps {rn-sae}, %zmm1, %zmm14, %zmm0{%k3} - -/* Go to special inputs processing branch */ - jne L(SPECIAL_VALUES_BRANCH) - # LOE rbx r12 r13 r14 r15 k0 zmm0 zmm8 - -/* Restore registers - * and exit the function - */ + /* Restore registers + * and exit the function + */ L(EXIT): - movq %rbp, %rsp - popq %rbp - cfi_def_cfa(7, 8) - cfi_restore(6) - ret - cfi_def_cfa(6, 16) - cfi_offset(6, -16) - -/* Branch to process - * special inputs - */ + movq %rbp, %rsp + popq %rbp + cfi_def_cfa(7, 8) + cfi_restore(6) + ret + cfi_def_cfa(6, 16) + cfi_offset(6, -16) + + /* Branch to process + * special inputs + */ L(SPECIAL_VALUES_BRANCH): - vmovups %zmm8, 64(%rsp) - vmovups %zmm0, 128(%rsp) - # LOE rbx r12 r13 r14 r15 k0 zmm0 - - xorl %eax, %eax - # LOE rbx r12 r13 r14 r15 eax k0 - - vzeroupper - movq %r12, 16(%rsp) - /* DW_CFA_expression: r12 (r12) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -176; DW_OP_plus) */ - .cfi_escape 0x10, 0x0c, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x50, 0xff, 0xff, 0xff, 0x22 - movl %eax, %r12d - movq %r13, 8(%rsp) - /* DW_CFA_expression: r13 (r13) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -184; DW_OP_plus) */ - .cfi_escape 0x10, 0x0d, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x48, 0xff, 0xff, 0xff, 0x22 - kmovd %k0, %r13d - movq %r14, (%rsp) - /* DW_CFA_expression: r14 (r14) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -192; DW_OP_plus) */ - .cfi_escape 0x10, 0x0e, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x40, 0xff, 0xff, 0xff, 0x22 - # LOE rbx r15 r12d r13d - -/* Range mask - * bits check - */ + vmovups %zmm8, 64(%rsp) + vmovups %zmm0, 128(%rsp) + # LOE rbx r12 r13 r14 r15 k0 zmm0 + + xorl %eax, %eax + # LOE rbx r12 r13 r14 r15 eax k0 + + vzeroupper + movq %r12, 16(%rsp) + /* DW_CFA_expression: r12 (r12) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -176; DW_OP_plus) */ + .cfi_escape 0x10, 0x0c, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x50, 0xff, 0xff, 0xff, 0x22 + movl %eax, %r12d + movq %r13, 8(%rsp) + /* DW_CFA_expression: r13 (r13) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -184; DW_OP_plus) */ + .cfi_escape 0x10, 0x0d, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x48, 0xff, 0xff, 0xff, 0x22 + kmovd %k0, %r13d + movq %r14, (%rsp) + /* DW_CFA_expression: r14 (r14) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -192; DW_OP_plus) */ + .cfi_escape 0x10, 0x0e, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x40, 0xff, 0xff, 0xff, 0x22 + # LOE rbx r15 r12d r13d + + /* Range mask + * bits check + */ L(RANGEMASK_CHECK): - btl %r12d, %r13d + btl %r12d, %r13d -/* Call scalar math function */ - jc L(SCALAR_MATH_CALL) - # LOE rbx r15 r12d r13d + /* Call scalar math function */ + jc L(SCALAR_MATH_CALL) + # LOE rbx r15 r12d r13d -/* Special inputs - * processing loop - */ + /* Special inputs + * processing loop + */ L(SPECIAL_VALUES_LOOP): - incl %r12d - cmpl $16, %r12d - -/* Check bits in range mask */ - jl L(RANGEMASK_CHECK) - # LOE rbx r15 r12d r13d - - movq 16(%rsp), %r12 - cfi_restore(12) - movq 8(%rsp), %r13 - cfi_restore(13) - movq (%rsp), %r14 - cfi_restore(14) - vmovups 128(%rsp), %zmm0 - -/* Go to exit */ - jmp L(EXIT) - /* DW_CFA_expression: r12 (r12) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -176; DW_OP_plus) */ - .cfi_escape 0x10, 0x0c, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x50, 0xff, 0xff, 0xff, 0x22 - /* DW_CFA_expression: r13 (r13) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -184; DW_OP_plus) */ - .cfi_escape 0x10, 0x0d, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x48, 0xff, 0xff, 0xff, 0x22 - /* DW_CFA_expression: r14 (r14) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -192; DW_OP_plus) */ - .cfi_escape 0x10, 0x0e, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x40, 0xff, 0xff, 0xff, 0x22 - # LOE rbx r12 r13 r14 r15 zmm0 - -/* Scalar math fucntion call - * to process special input - */ + incl %r12d + cmpl $16, %r12d + + /* Check bits in range mask */ + jl L(RANGEMASK_CHECK) + # LOE rbx r15 r12d r13d + + movq 16(%rsp), %r12 + cfi_restore(12) + movq 8(%rsp), %r13 + cfi_restore(13) + movq (%rsp), %r14 + cfi_restore(14) + vmovups 128(%rsp), %zmm0 + + /* Go to exit */ + jmp L(EXIT) + /* DW_CFA_expression: r12 (r12) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -176; DW_OP_plus) */ + .cfi_escape 0x10, 0x0c, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x50, 0xff, 0xff, 0xff, 0x22 + /* DW_CFA_expression: r13 (r13) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -184; DW_OP_plus) */ + .cfi_escape 0x10, 0x0d, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x48, 0xff, 0xff, 0xff, 0x22 + /* DW_CFA_expression: r14 (r14) (DW_OP_lit8; DW_OP_minus; DW_OP_const4s: -64; DW_OP_and; DW_OP_const4s: -192; DW_OP_plus) */ + .cfi_escape 0x10, 0x0e, 0x0e, 0x38, 0x1c, 0x0d, 0xc0, 0xff, 0xff, 0xff, 0x1a, 0x0d, 0x40, 0xff, 0xff, 0xff, 0x22 + # LOE rbx r12 r13 r14 r15 zmm0 + + /* Scalar math fucntion call + * to process special input + */ L(SCALAR_MATH_CALL): - movl %r12d, %r14d - movss 64(%rsp,%r14,4), %xmm0 - call acoshf@PLT - # LOE rbx r14 r15 r12d r13d xmm0 + movl %r12d, %r14d + movss 64(%rsp, %r14, 4), %xmm0 + call acoshf@PLT + # LOE rbx r14 r15 r12d r13d xmm0 - movss %xmm0, 128(%rsp,%r14,4) + movss %xmm0, 128(%rsp, %r14, 4) -/* Process special inputs in loop */ - jmp L(SPECIAL_VALUES_LOOP) - # LOE rbx r15 r12d r13d + /* Process special inputs in loop */ + jmp L(SPECIAL_VALUES_LOOP) + # LOE rbx r15 r12d r13d END(_ZGVeN16v_acoshf_skx) - .section .rodata, "a" - .align 64 + .section .rodata, "a" + .align 64 #ifdef __svml_sacosh_data_internal_avx512_typedef typedef unsigned int VUINT32; typedef struct { - __declspec(align(64)) VUINT32 Log_tbl_H[32][1]; - __declspec(align(64)) VUINT32 Log_tbl_L[32][1]; - __declspec(align(64)) VUINT32 One[16][1]; - __declspec(align(64)) VUINT32 SmallThreshold[16][1]; - __declspec(align(64)) VUINT32 Threshold[16][1]; - __declspec(align(64)) VUINT32 LargeThreshold[16][1]; - __declspec(align(64)) VUINT32 ca1[16][1]; - __declspec(align(64)) VUINT32 c2s[16][1]; - __declspec(align(64)) VUINT32 c1s[16][1]; - __declspec(align(64)) VUINT32 AddB5[16][1]; - __declspec(align(64)) VUINT32 RcpBitMask[16][1]; - __declspec(align(64)) VUINT32 OneEighth[16][1]; - __declspec(align(64)) VUINT32 Four[16][1]; - __declspec(align(64)) VUINT32 poly_coeff3[16][1]; - __declspec(align(64)) VUINT32 poly_coeff2[16][1]; - __declspec(align(64)) VUINT32 poly_coeff1[16][1]; - __declspec(align(64)) VUINT32 L2H[16][1]; - __declspec(align(64)) VUINT32 L2L[16][1]; - } __svml_sacosh_data_internal_avx512; + __declspec(align(64)) VUINT32 Log_tbl_H[32][1]; + __declspec(align(64)) VUINT32 Log_tbl_L[32][1]; + __declspec(align(64)) VUINT32 One[16][1]; + __declspec(align(64)) VUINT32 SmallThreshold[16][1]; + __declspec(align(64)) VUINT32 Threshold[16][1]; + __declspec(align(64)) VUINT32 LargeThreshold[16][1]; + __declspec(align(64)) VUINT32 ca1[16][1]; + __declspec(align(64)) VUINT32 c2s[16][1]; + __declspec(align(64)) VUINT32 c1s[16][1]; + __declspec(align(64)) VUINT32 AddB5[16][1]; + __declspec(align(64)) VUINT32 RcpBitMask[16][1]; + __declspec(align(64)) VUINT32 OneEighth[16][1]; + __declspec(align(64)) VUINT32 Four[16][1]; + __declspec(align(64)) VUINT32 poly_coeff3[16][1]; + __declspec(align(64)) VUINT32 poly_coeff2[16][1]; + __declspec(align(64)) VUINT32 poly_coeff1[16][1]; + __declspec(align(64)) VUINT32 L2H[16][1]; + __declspec(align(64)) VUINT32 L2L[16][1]; +} __svml_sacosh_data_internal_avx512; #endif __svml_sacosh_data_internal_avx512: - /*== Log_tbl_H ==*/ - .long 0x00000000 - .long 0xbcfc0000 - .long 0xbd788000 - .long 0xbdb78000 - .long 0xbdf14000 - .long 0xbe14a000 - .long 0xbe300000 - .long 0xbe4aa000 - .long 0xbe648000 - .long 0xbe7dc000 - .long 0xbe8b4000 - .long 0xbe974000 - .long 0xbea31000 - .long 0xbeae9000 - .long 0xbeb9d000 - .long 0xbec4d000 - .long 0xbecfa000 - .long 0xbeda2000 - .long 0xbee48000 - .long 0xbeeea000 - .long 0xbef89000 - .long 0xbf012800 - .long 0xbf05f000 - .long 0xbf0aa800 - .long 0xbf0f4000 - .long 0xbf13c800 - .long 0xbf184000 - .long 0xbf1ca000 - .long 0xbf20f000 - .long 0xbf252800 - .long 0xbf295000 - .long 0xbf2d6800 - /*== Log_tbl_L ==*/ - .align 64 - .long 0x80000000 - .long 0xb726c39e - .long 0x3839e7fe - .long 0xb7528ae5 - .long 0x377891d5 - .long 0xb8297c10 - .long 0x37cf8f58 - .long 0x3852b186 - .long 0x35838656 - .long 0xb80c36af - .long 0x38235454 - .long 0xb862bae1 - .long 0x37e87bc7 - .long 0x37848150 - .long 0x37202511 - .long 0xb74e1b05 - .long 0x385c1340 - .long 0xb8777bcd - .long 0x36038656 - .long 0xb7d40984 - .long 0xb80f5faf - .long 0xb8254b4c - .long 0xb865c84a - .long 0x37f0b42d - .long 0xb83ebce1 - .long 0xb83c2513 - .long 0x37a332c4 - .long 0x3779654f - .long 0x38602f73 - .long 0x367449f8 - .long 0xb7b4996f - .long 0xb800986b - /*== One ==*/ - .align 64 - .long 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000 - /*== SmallThreshold ==*/ - .align 64 - .long 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000 - /*== Threshold ==*/ - .align 64 - .long 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000 - /*== LargeThreshold ==*/ - .align 64 - .long 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff - /*== ca1 ==*/ - .align 64 - .long 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE - /*== c2s ==*/ - .align 64 - .long 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000 - /*== c1s ==*/ - .align 64 - .long 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000 - /*== AddB5 ==*/ - .align 64 - .long 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000 - /*== RcpBitMask ==*/ - .align 64 - .long 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000 - /*==OneEighth ==*/ - .align 64 - .long 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000 - /*== Four ==*/ - .align 64 - .long 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000 - /*== poly_coeff3 ==*/ - .align 64 - .long 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810 - /*== poly_coeff2 ==*/ - .align 64 - .long 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e - /*== poly_coeff1 ==*/ - .align 64 - .long 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000 - /*== L2H = log(2)_high ==*/ - .align 64 - .long 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000 - /*== L2L = log(2)_low ==*/ - .align 64 - .long 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4 - .align 64 - .type __svml_sacosh_data_internal_avx512,@object - .size __svml_sacosh_data_internal_avx512,.-__svml_sacosh_data_internal_avx512 + /* Log_tbl_H */ + .long 0x00000000 + .long 0xbcfc0000 + .long 0xbd788000 + .long 0xbdb78000 + .long 0xbdf14000 + .long 0xbe14a000 + .long 0xbe300000 + .long 0xbe4aa000 + .long 0xbe648000 + .long 0xbe7dc000 + .long 0xbe8b4000 + .long 0xbe974000 + .long 0xbea31000 + .long 0xbeae9000 + .long 0xbeb9d000 + .long 0xbec4d000 + .long 0xbecfa000 + .long 0xbeda2000 + .long 0xbee48000 + .long 0xbeeea000 + .long 0xbef89000 + .long 0xbf012800 + .long 0xbf05f000 + .long 0xbf0aa800 + .long 0xbf0f4000 + .long 0xbf13c800 + .long 0xbf184000 + .long 0xbf1ca000 + .long 0xbf20f000 + .long 0xbf252800 + .long 0xbf295000 + .long 0xbf2d6800 + /* Log_tbl_L */ + .align 64 + .long 0x80000000 + .long 0xb726c39e + .long 0x3839e7fe + .long 0xb7528ae5 + .long 0x377891d5 + .long 0xb8297c10 + .long 0x37cf8f58 + .long 0x3852b186 + .long 0x35838656 + .long 0xb80c36af + .long 0x38235454 + .long 0xb862bae1 + .long 0x37e87bc7 + .long 0x37848150 + .long 0x37202511 + .long 0xb74e1b05 + .long 0x385c1340 + .long 0xb8777bcd + .long 0x36038656 + .long 0xb7d40984 + .long 0xb80f5faf + .long 0xb8254b4c + .long 0xb865c84a + .long 0x37f0b42d + .long 0xb83ebce1 + .long 0xb83c2513 + .long 0x37a332c4 + .long 0x3779654f + .long 0x38602f73 + .long 0x367449f8 + .long 0xb7b4996f + .long 0xb800986b + /* One */ + .align 64 + .long 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000 + /* SmallThreshold */ + .align 64 + .long 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000, 0x39800000 + /* Threshold */ + .align 64 + .long 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000, 0x5f000000 + /* LargeThreshold */ + .align 64 + .long 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff, 0x7f7fffff + /* ca1 */ + .align 64 + .long 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE, 0xbe2AA5DE + /* c2s */ + .align 64 + .long 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000, 0x3ec00000 + /* c1s */ + .align 64 + .long 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000 + /* AddB5 */ + .align 64 + .long 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000, 0x00020000 + /* RcpBitMask */ + .align 64 + .long 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000, 0xfffc0000 + /* OneEighth */ + .align 64 + .long 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000, 0x3e000000 + /* Four */ + .align 64 + .long 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000, 0x40800000 + /* poly_coeff3 */ + .align 64 + .long 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810, 0xbe800810 + /* poly_coeff2 */ + .align 64 + .long 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e, 0x3eaab11e + /* poly_coeff1 */ + .align 64 + .long 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000, 0xbf000000 + /* L2H = log(2)_high */ + .align 64 + .long 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000, 0x3f317000 + /* L2L = log(2)_low */ + .align 64 + .long 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4, 0x3805fdf4 + .align 64 + .type __svml_sacosh_data_internal_avx512, @object + .size __svml_sacosh_data_internal_avx512, .-__svml_sacosh_data_internal_avx512 -- 2.7.4