x86_64: Fix svml_s_asinhf16_core_avx512.S code formatting
authorSunil K Pandey <skpgkp2@gmail.com>
Mon, 7 Mar 2022 18:47:09 +0000 (10:47 -0800)
committerSunil K Pandey <skpgkp2@gmail.com>
Tue, 8 Mar 2022 05:14:09 +0000 (21:14 -0800)
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 <goldstein.w.n@gmail.com>
sysdeps/x86_64/fpu/multiarch/svml_s_asinhf16_core_avx512.S

index 3cd80db..6b569ec 100644 (file)
 
 /* Offsets for data table __svml_sasinh_data_internal_avx512
  */
-#define Log_tbl_H                      0
-#define Log_tbl_L                      128
-#define One                            256
-#define AbsMask                        320
-#define SmallThreshold                 384
-#define Threshold                      448
-#define LargeThreshold                 512
-#define ca1                            576
-#define c2s                            640
-#define c1s                            704
-#define AddB5                          768
-#define RcpBitMask                     832
-#define OneEighth                      896
-#define Four                           960
-#define poly_coeff3                    1024
-#define poly_coeff2                    1088
-#define poly_coeff1                    1152
-#define L2H                            1216
-#define L2L                            1280
+#define Log_tbl_H                      0
+#define Log_tbl_L                      128
+#define One                            256
+#define AbsMask                                320
+#define SmallThreshold                 384
+#define Threshold                      448
+#define LargeThreshold                 512
+#define ca1                            576
+#define c2s                            640
+#define c1s                            704
+#define AddB5                          768
+#define RcpBitMask                     832
+#define OneEighth                      896
+#define Four                           960
+#define poly_coeff3                    1024
+#define poly_coeff2                    1088
+#define poly_coeff1                    1152
+#define L2H                            1216
+#define L2L                            1280
 
 #include <sysdep.h>
 
-        .text
-       .section .text.exex512,"ax",@progbits
+       .section .text.exex512, "ax", @progbits
 ENTRY(_ZGVeN16v_asinhf_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
-        vmovaps   %zmm0, %zmm10
+       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
+       vmovaps %zmm0, %zmm10
 
-/* x^2 */
-        vmulps    {rn-sae}, %zmm10, %zmm10, %zmm0
-        vmovups   One+__svml_sasinh_data_internal_avx512(%rip), %zmm2
+       /* x^2 */
+       vmulps  {rn-sae}, %zmm10, %zmm10, %zmm0
+       vmovups One+__svml_sasinh_data_internal_avx512(%rip), %zmm2
 
-/* polynomial computation for small inputs */
-        vmovups   ca1+__svml_sasinh_data_internal_avx512(%rip), %zmm1
+       /* polynomial computation for small inputs */
+       vmovups ca1+__svml_sasinh_data_internal_avx512(%rip), %zmm1
 
-/* not a very small input ? */
-        vmovups   SmallThreshold+__svml_sasinh_data_internal_avx512(%rip), %zmm11
+       /* not a very small input ? */
+       vmovups SmallThreshold+__svml_sasinh_data_internal_avx512(%rip), %zmm11
 
-/* 1+x^2 */
-        vaddps    {rn-sae}, %zmm2, %zmm0, %zmm7
+       /* 1+x^2 */
+       vaddps  {rn-sae}, %zmm2, %zmm0, %zmm7
 
-/* |input| */
-        vandps    AbsMask+__svml_sasinh_data_internal_avx512(%rip), %zmm10, %zmm12
+       /* |input| */
+       vandps  AbsMask+__svml_sasinh_data_internal_avx512(%rip), %zmm10, %zmm12
 
-/* A=max(x^2, 1); */
-        vmaxps    {sae}, %zmm0, %zmm2, %zmm14
-        vrsqrt14ps %zmm7, %zmm8
+       /* A=max(x^2, 1); */
+       vmaxps  {sae}, %zmm0, %zmm2, %zmm14
+       vrsqrt14ps %zmm7, %zmm8
 
-/* B=min(x^2, 1); */
-        vminps    {sae}, %zmm0, %zmm2, %zmm15
-        vcmpps    $21, {sae}, %zmm11, %zmm12, %k2
+       /* B=min(x^2, 1); */
+       vminps  {sae}, %zmm0, %zmm2, %zmm15
+       vcmpps  $21, {sae}, %zmm11, %zmm12, %k2
 
-/* B_high */
-        vsubps    {rn-sae}, %zmm14, %zmm7, %zmm9
+       /* B_high */
+       vsubps  {rn-sae}, %zmm14, %zmm7, %zmm9
 
-/* sign bit */
-        vxorps    %zmm10, %zmm12, %zmm13
+       /* sign bit */
+       vxorps  %zmm10, %zmm12, %zmm13
 
-/* Sh ~sqrt(1+x^2) */
-        vmulps    {rn-sae}, %zmm8, %zmm7, %zmm6
-        vmovups   LargeThreshold+__svml_sasinh_data_internal_avx512(%rip), %zmm14
+       /* Sh ~sqrt(1+x^2) */
+       vmulps  {rn-sae}, %zmm8, %zmm7, %zmm6
+       vmovups LargeThreshold+__svml_sasinh_data_internal_avx512(%rip), %zmm14
 
-/* B_low */
-        vsubps    {rn-sae}, %zmm9, %zmm15, %zmm3
+       /* B_low */
+       vsubps  {rn-sae}, %zmm9, %zmm15, %zmm3
 
-/* Sh+x */
-        vaddps    {rn-sae}, %zmm12, %zmm6, %zmm15
+       /* Sh+x */
+       vaddps  {rn-sae}, %zmm12, %zmm6, %zmm15
 
-/* (Yh*R0)_low */
-        vfmsub213ps {rn-sae}, %zmm6, %zmm8, %zmm7
-        vmulps    {rn-sae}, %zmm1, %zmm0, %zmm9
-        vcmpps    $22, {sae}, %zmm14, %zmm12, %k0
-        vmovups   c1s+__svml_sasinh_data_internal_avx512(%rip), %zmm1
+       /* (Yh*R0)_low */
+       vfmsub213ps {rn-sae}, %zmm6, %zmm8, %zmm7
+       vmulps  {rn-sae}, %zmm1, %zmm0, %zmm9
+       vcmpps  $22, {sae}, %zmm14, %zmm12, %k0
+       vmovups c1s+__svml_sasinh_data_internal_avx512(%rip), %zmm1
 
-/* polynomial computation for small inputs */
-        vfmadd213ps {rn-sae}, %zmm12, %zmm12, %zmm9
-        kmovw     %k0, %edx
+       /* polynomial computation for small inputs */
+       vfmadd213ps {rn-sae}, %zmm12, %zmm12, %zmm9
+       kmovw   %k0, %edx
 
-/* (x^2)_low */
-        vmovaps   %zmm10, %zmm4
-        vfmsub213ps {rn-sae}, %zmm0, %zmm10, %zmm4
+       /* (x^2)_low */
+       vmovaps %zmm10, %zmm4
+       vfmsub213ps {rn-sae}, %zmm0, %zmm10, %zmm4
 
-/* Yl = (x^2)_low + B_low */
-        vaddps    {rn-sae}, %zmm4, %zmm3, %zmm5
+       /* Yl = (x^2)_low + B_low */
+       vaddps  {rn-sae}, %zmm4, %zmm3, %zmm5
 
-/* rel. error term: Eh=1-Sh*R0 */
-        vmovaps   %zmm2, %zmm0
-        vfnmadd231ps {rn-sae}, %zmm6, %zmm8, %zmm0
+       /* rel. error term: Eh=1-Sh*R0 */
+       vmovaps %zmm2, %zmm0
+       vfnmadd231ps {rn-sae}, %zmm6, %zmm8, %zmm0
 
-/* Sl = (Yh*R0)_low+(R0*Yl) */
-        vfmadd213ps {rn-sae}, %zmm7, %zmm8, %zmm5
+       /* Sl = (Yh*R0)_low+(R0*Yl) */
+       vfmadd213ps {rn-sae}, %zmm7, %zmm8, %zmm5
 
-/* very large inputs ? */
-        vmovups   Threshold+__svml_sasinh_data_internal_avx512(%rip), %zmm7
+       /* very large inputs ? */
+       vmovups Threshold+__svml_sasinh_data_internal_avx512(%rip), %zmm7
 
-/* rel. error term: Eh=(1-Sh*R0)-Sl*R0 */
-        vfnmadd231ps {rn-sae}, %zmm5, %zmm8, %zmm0
+       /* rel. error term: Eh=(1-Sh*R0)-Sl*R0 */
+       vfnmadd231ps {rn-sae}, %zmm5, %zmm8, %zmm0
 
-/* sqrt(1+x^2) ~ Sh + Sl + Sh*Eh*poly_s */
-        vmovups   c2s+__svml_sasinh_data_internal_avx512(%rip), %zmm8
-        vcmpps    $21, {sae}, %zmm7, %zmm12, %k1
+       /* sqrt(1+x^2) ~ Sh + Sl + Sh*Eh*poly_s */
+       vmovups c2s+__svml_sasinh_data_internal_avx512(%rip), %zmm8
+       vcmpps  $21, {sae}, %zmm7, %zmm12, %k1
 
-/* Sh*Eh */
-        vmulps    {rn-sae}, %zmm0, %zmm6, %zmm4
-        vfmadd231ps {rn-sae}, %zmm0, %zmm8, %zmm1
+       /* Sh*Eh */
+       vmulps  {rn-sae}, %zmm0, %zmm6, %zmm4
+       vfmadd231ps {rn-sae}, %zmm0, %zmm8, %zmm1
 
-/* Sl + Sh*Eh*poly_s */
-        vfmadd213ps {rn-sae}, %zmm5, %zmm1, %zmm4
+       /* Sl + Sh*Eh*poly_s */
+       vfmadd213ps {rn-sae}, %zmm5, %zmm1, %zmm4
 
-/* Xh */
-        vsubps    {rn-sae}, %zmm6, %zmm15, %zmm5
+       /* Xh */
+       vsubps  {rn-sae}, %zmm6, %zmm15, %zmm5
 
-/* fixup for very large inputs */
-        vmovups   OneEighth+__svml_sasinh_data_internal_avx512(%rip), %zmm6
+       /* fixup for very large inputs */
+       vmovups OneEighth+__svml_sasinh_data_internal_avx512(%rip), %zmm6
 
-/* Xin0+Sl+Sh*Eh*poly_s ~ x+sqrt(1+x^2) */
-        vaddps    {rn-sae}, %zmm4, %zmm15, %zmm3
+       /* Xin0+Sl+Sh*Eh*poly_s ~ x+sqrt(1+x^2) */
+       vaddps  {rn-sae}, %zmm4, %zmm15, %zmm3
 
-/* Xl */
-        vsubps    {rn-sae}, %zmm5, %zmm12, %zmm5
+       /* Xl */
+       vsubps  {rn-sae}, %zmm5, %zmm12, %zmm5
 
-/* Sl_high */
-        vsubps    {rn-sae}, %zmm15, %zmm3, %zmm0
-        vmulps    {rn-sae}, %zmm6, %zmm12, %zmm3{%k1}
+       /* Sl_high */
+       vsubps  {rn-sae}, %zmm15, %zmm3, %zmm0
+       vmulps  {rn-sae}, %zmm6, %zmm12, %zmm3{%k1}
 
-/* -K*L2H + Th */
-        vmovups   L2H+__svml_sasinh_data_internal_avx512(%rip), %zmm15
+       /* -K*L2H + Th */
+       vmovups L2H+__svml_sasinh_data_internal_avx512(%rip), %zmm15
 
-/* Sl_l */
-        vsubps    {rn-sae}, %zmm0, %zmm4, %zmm1
-        vrcp14ps  %zmm3, %zmm6
+       /* Sl_l */
+       vsubps  {rn-sae}, %zmm0, %zmm4, %zmm1
+       vrcp14ps %zmm3, %zmm6
 
-/* Table lookups */
-        vmovups   __svml_sasinh_data_internal_avx512(%rip), %zmm0
+       /* Table lookups */
+       vmovups __svml_sasinh_data_internal_avx512(%rip), %zmm0
 
-/* Xin_low */
-        vaddps    {rn-sae}, %zmm5, %zmm1, %zmm7
+       /* Xin_low */
+       vaddps  {rn-sae}, %zmm5, %zmm1, %zmm7
 
-/* round reciprocal to 1+4b mantissas */
-        vpaddd    AddB5+__svml_sasinh_data_internal_avx512(%rip), %zmm6, %zmm4
-        vmovups   poly_coeff1+__svml_sasinh_data_internal_avx512(%rip), %zmm5
-        vandps    RcpBitMask+__svml_sasinh_data_internal_avx512(%rip), %zmm4, %zmm8
+       /* round reciprocal to 1+4b mantissas */
+       vpaddd  AddB5+__svml_sasinh_data_internal_avx512(%rip), %zmm6, %zmm4
+       vmovups poly_coeff1+__svml_sasinh_data_internal_avx512(%rip), %zmm5
+       vandps  RcpBitMask+__svml_sasinh_data_internal_avx512(%rip), %zmm4, %zmm8
 
-/* fixup for very large inputs */
-        vxorps    %zmm7, %zmm7, %zmm7{%k1}
+       /* fixup for very large inputs */
+       vxorps  %zmm7, %zmm7, %zmm7{%k1}
 
-/* polynomial */
-        vmovups   poly_coeff3+__svml_sasinh_data_internal_avx512(%rip), %zmm4
+       /* polynomial */
+       vmovups poly_coeff3+__svml_sasinh_data_internal_avx512(%rip), %zmm4
 
-/* reduced argument for log(): (Rcp*Xin-1)+Rcp*Xin_low */
-        vfmsub231ps {rn-sae}, %zmm8, %zmm3, %zmm2
-        vmovups   Four+__svml_sasinh_data_internal_avx512(%rip), %zmm3
+       /* reduced argument for log(): (Rcp*Xin-1)+Rcp*Xin_low */
+       vfmsub231ps {rn-sae}, %zmm8, %zmm3, %zmm2
+       vmovups Four+__svml_sasinh_data_internal_avx512(%rip), %zmm3
 
-/* exponents */
-        vgetexpps {sae}, %zmm8, %zmm1
+       /* exponents */
+       vgetexpps {sae}, %zmm8, %zmm1
 
-/* Prepare table index */
-        vpsrld    $18, %zmm8, %zmm14
-        vfmadd231ps {rn-sae}, %zmm8, %zmm7, %zmm2
-        vmovups   poly_coeff2+__svml_sasinh_data_internal_avx512(%rip), %zmm7
-        vsubps    {rn-sae}, %zmm3, %zmm1, %zmm1{%k1}
-        vpermt2ps Log_tbl_H+64+__svml_sasinh_data_internal_avx512(%rip), %zmm14, %zmm0
-        vmovups   Log_tbl_L+__svml_sasinh_data_internal_avx512(%rip), %zmm3
-        vfmadd231ps {rn-sae}, %zmm2, %zmm4, %zmm7
-        vfnmadd231ps {rn-sae}, %zmm1, %zmm15, %zmm0
+       /* Prepare table index */
+       vpsrld  $18, %zmm8, %zmm14
+       vfmadd231ps {rn-sae}, %zmm8, %zmm7, %zmm2
+       vmovups poly_coeff2+__svml_sasinh_data_internal_avx512(%rip), %zmm7
+       vsubps  {rn-sae}, %zmm3, %zmm1, %zmm1{%k1}
+       vpermt2ps Log_tbl_H+64+__svml_sasinh_data_internal_avx512(%rip), %zmm14, %zmm0
+       vmovups Log_tbl_L+__svml_sasinh_data_internal_avx512(%rip), %zmm3
+       vfmadd231ps {rn-sae}, %zmm2, %zmm4, %zmm7
+       vfnmadd231ps {rn-sae}, %zmm1, %zmm15, %zmm0
 
-/* R^2 */
-        vmulps    {rn-sae}, %zmm2, %zmm2, %zmm6
-        vfmadd213ps {rn-sae}, %zmm5, %zmm2, %zmm7
-        vpermt2ps Log_tbl_L+64+__svml_sasinh_data_internal_avx512(%rip), %zmm14, %zmm3
+       /* R^2 */
+       vmulps  {rn-sae}, %zmm2, %zmm2, %zmm6
+       vfmadd213ps {rn-sae}, %zmm5, %zmm2, %zmm7
+       vpermt2ps Log_tbl_L+64+__svml_sasinh_data_internal_avx512(%rip), %zmm14, %zmm3
 
-/* -K*L2L + Tl */
-        vmovups   L2L+__svml_sasinh_data_internal_avx512(%rip), %zmm14
-        vfnmadd213ps {rn-sae}, %zmm3, %zmm14, %zmm1
+       /* -K*L2L + Tl */
+       vmovups L2L+__svml_sasinh_data_internal_avx512(%rip), %zmm14
+       vfnmadd213ps {rn-sae}, %zmm3, %zmm14, %zmm1
 
-/* Tl + R^2*Poly */
-        vfmadd213ps {rn-sae}, %zmm1, %zmm6, %zmm7
+       /* Tl + R^2*Poly */
+       vfmadd213ps {rn-sae}, %zmm1, %zmm6, %zmm7
 
-/* R+Tl + R^2*Poly */
-        vaddps    {rn-sae}, %zmm2, %zmm7, %zmm2
-        vaddps    {rn-sae}, %zmm2, %zmm0, %zmm9{%k2}
-        vxorps    %zmm13, %zmm9, %zmm0
-        testl     %edx, %edx
+       /* R+Tl + R^2*Poly */
+       vaddps  {rn-sae}, %zmm2, %zmm7, %zmm2
+       vaddps  {rn-sae}, %zmm2, %zmm0, %zmm9{%k2}
+       vxorps  %zmm13, %zmm9, %zmm0
+       testl   %edx, %edx
 
-/* Go to special inputs processing branch */
-        jne       L(SPECIAL_VALUES_BRANCH)
-                                # LOE rbx r12 r13 r14 r15 edx zmm0 zmm10
+       /* Go to special inputs processing branch */
+       jne     L(SPECIAL_VALUES_BRANCH)
+       # LOE rbx r12 r13 r14 r15 edx zmm0 zmm10
 
-/* 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   %zmm10, 64(%rsp)
-        vmovups   %zmm0, 128(%rsp)
-                                # LOE rbx r12 r13 r14 r15 edx zmm0
-
-        xorl      %eax, %eax
-                                # LOE rbx r12 r13 r14 r15 eax edx
-
-        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
-        movl      %edx, %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 %zmm10, 64(%rsp)
+       vmovups %zmm0, 128(%rsp)
+       # LOE rbx r12 r13 r14 r15 edx zmm0
+
+       xorl    %eax, %eax
+       # LOE rbx r12 r13 r14 r15 eax edx
+
+       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
+       movl    %edx, %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      asinhf@PLT
-                                # LOE rbx r14 r15 r12d r13d xmm0
+       movl    %r12d, %r14d
+       movss   64(%rsp, %r14, 4), %xmm0
+       call    asinhf@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_asinhf_skx)
 
-        .section .rodata, "a"
-        .align 64
+       .section .rodata, "a"
+       .align  64
 
 #ifdef __svml_sasinh_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 AbsMask[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_sasinh_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 AbsMask[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_sasinh_data_internal_avx512;
 #endif
 __svml_sasinh_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
-        /*== AbsMask ==*/
-        .align 64
-        .long 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff
-        /*== SmallThreshold ==*/
-        .align 64
-        .long 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000
-        /*== 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_sasinh_data_internal_avx512,@object
-        .size  __svml_sasinh_data_internal_avx512,.-__svml_sasinh_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
+       /* AbsMask */
+       .align  64
+       .long   0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff
+       /* SmallThreshold */
+       .align  64
+       .long   0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000, 0x3c800000
+       /* 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_sasinh_data_internal_avx512, @object
+       .size   __svml_sasinh_data_internal_avx512, .-__svml_sasinh_data_internal_avx512