@@ -28,225 +28,224 @@
/* Offsets for data table __svml_sasin_data_internal
*/
-#define AbsMask 0
-#define OneHalf 16
-#define SmallNorm 32
-#define One 48
-#define Two 64
-#define sqrt_coeff 80
-#define poly_coeff 112
-#define Pi2H 192
+#define AbsMask 0
+#define OneHalf 16
+#define SmallNorm 32
+#define One 48
+#define Two 64
+#define sqrt_coeff 80
+#define poly_coeff 112
+#define Pi2H 192
#include <sysdep.h>
- .text
- .section .text.sse4,"ax",@progbits
+ .section .text.sse4, "ax", @progbits
ENTRY(_ZGVbN4v_asinf_sse4)
- subq $72, %rsp
- cfi_def_cfa_offset(80)
- movaps %xmm0, %xmm2
- movups __svml_sasin_data_internal(%rip), %xmm1
- movups OneHalf+__svml_sasin_data_internal(%rip), %xmm5
-
-/* x = |arg| */
- movaps %xmm1, %xmm0
- andps %xmm2, %xmm0
-
-/* Y = 0.5 - 0.5*x */
- movaps %xmm5, %xmm3
- mulps %xmm0, %xmm3
- movaps %xmm5, %xmm8
-
-/* x^2 */
- movaps %xmm0, %xmm14
- movaps %xmm0, %xmm15
- mulps %xmm0, %xmm14
- subps %xmm3, %xmm8
- cmpnltps %xmm5, %xmm15
-
-/* SQ ~ -2*sqrt(Y) */
- rsqrtps %xmm8, %xmm6
- minps %xmm8, %xmm14
- movaps %xmm8, %xmm9
- movaps %xmm14, %xmm10
- cmpltps SmallNorm+__svml_sasin_data_internal(%rip), %xmm9
- mulps %xmm14, %xmm10
- addps %xmm8, %xmm8
- andnps %xmm6, %xmm9
- movaps %xmm15, %xmm3
- movaps %xmm9, %xmm7
- andnps %xmm0, %xmm3
- mulps %xmm9, %xmm7
- andnps %xmm2, %xmm1
- mulps %xmm8, %xmm9
- mulps %xmm7, %xmm8
-
-/* polynomial */
- movups poly_coeff+__svml_sasin_data_internal(%rip), %xmm11
- mulps %xmm14, %xmm11
- subps Two+__svml_sasin_data_internal(%rip), %xmm8
- movups poly_coeff+32+__svml_sasin_data_internal(%rip), %xmm12
- mulps %xmm14, %xmm12
- addps poly_coeff+16+__svml_sasin_data_internal(%rip), %xmm11
- mulps %xmm10, %xmm11
- addps poly_coeff+48+__svml_sasin_data_internal(%rip), %xmm12
- movups sqrt_coeff+__svml_sasin_data_internal(%rip), %xmm13
- addps %xmm11, %xmm12
- mulps %xmm8, %xmm13
- mulps %xmm9, %xmm8
- mulps %xmm14, %xmm12
- addps sqrt_coeff+16+__svml_sasin_data_internal(%rip), %xmm13
- addps poly_coeff+64+__svml_sasin_data_internal(%rip), %xmm12
- mulps %xmm8, %xmm13
- mulps %xmm12, %xmm14
- subps %xmm9, %xmm13
- andps %xmm15, %xmm13
- orps %xmm13, %xmm3
- mulps %xmm3, %xmm14
- movups One+__svml_sasin_data_internal(%rip), %xmm4
- addps %xmm14, %xmm3
- cmpltps %xmm0, %xmm4
- movups Pi2H+__svml_sasin_data_internal(%rip), %xmm0
- andps %xmm15, %xmm0
- movmskps %xmm4, %edx
- addps %xmm3, %xmm0
- pxor %xmm1, %xmm0
- testl %edx, %edx
-
-/* Go to special inputs processing branch */
- jne L(SPECIAL_VALUES_BRANCH)
- # LOE rbx rbp r12 r13 r14 r15 edx xmm0 xmm2
-
-/* Restore registers
- * and exit the function
- */
+ subq $72, %rsp
+ cfi_def_cfa_offset(80)
+ movaps %xmm0, %xmm2
+ movups __svml_sasin_data_internal(%rip), %xmm1
+ movups OneHalf+__svml_sasin_data_internal(%rip), %xmm5
+
+ /* x = |arg| */
+ movaps %xmm1, %xmm0
+ andps %xmm2, %xmm0
+
+ /* Y = 0.5 - 0.5*x */
+ movaps %xmm5, %xmm3
+ mulps %xmm0, %xmm3
+ movaps %xmm5, %xmm8
+
+ /* x^2 */
+ movaps %xmm0, %xmm14
+ movaps %xmm0, %xmm15
+ mulps %xmm0, %xmm14
+ subps %xmm3, %xmm8
+ cmpnltps %xmm5, %xmm15
+
+ /* SQ ~ -2*sqrt(Y) */
+ rsqrtps %xmm8, %xmm6
+ minps %xmm8, %xmm14
+ movaps %xmm8, %xmm9
+ movaps %xmm14, %xmm10
+ cmpltps SmallNorm+__svml_sasin_data_internal(%rip), %xmm9
+ mulps %xmm14, %xmm10
+ addps %xmm8, %xmm8
+ andnps %xmm6, %xmm9
+ movaps %xmm15, %xmm3
+ movaps %xmm9, %xmm7
+ andnps %xmm0, %xmm3
+ mulps %xmm9, %xmm7
+ andnps %xmm2, %xmm1
+ mulps %xmm8, %xmm9
+ mulps %xmm7, %xmm8
+
+ /* polynomial */
+ movups poly_coeff+__svml_sasin_data_internal(%rip), %xmm11
+ mulps %xmm14, %xmm11
+ subps Two+__svml_sasin_data_internal(%rip), %xmm8
+ movups poly_coeff+32+__svml_sasin_data_internal(%rip), %xmm12
+ mulps %xmm14, %xmm12
+ addps poly_coeff+16+__svml_sasin_data_internal(%rip), %xmm11
+ mulps %xmm10, %xmm11
+ addps poly_coeff+48+__svml_sasin_data_internal(%rip), %xmm12
+ movups sqrt_coeff+__svml_sasin_data_internal(%rip), %xmm13
+ addps %xmm11, %xmm12
+ mulps %xmm8, %xmm13
+ mulps %xmm9, %xmm8
+ mulps %xmm14, %xmm12
+ addps sqrt_coeff+16+__svml_sasin_data_internal(%rip), %xmm13
+ addps poly_coeff+64+__svml_sasin_data_internal(%rip), %xmm12
+ mulps %xmm8, %xmm13
+ mulps %xmm12, %xmm14
+ subps %xmm9, %xmm13
+ andps %xmm15, %xmm13
+ orps %xmm13, %xmm3
+ mulps %xmm3, %xmm14
+ movups One+__svml_sasin_data_internal(%rip), %xmm4
+ addps %xmm14, %xmm3
+ cmpltps %xmm0, %xmm4
+ movups Pi2H+__svml_sasin_data_internal(%rip), %xmm0
+ andps %xmm15, %xmm0
+ movmskps %xmm4, %edx
+ addps %xmm3, %xmm0
+ pxor %xmm1, %xmm0
+ testl %edx, %edx
+
+ /* Go to special inputs processing branch */
+ jne L(SPECIAL_VALUES_BRANCH)
+ # LOE rbx rbp r12 r13 r14 r15 edx xmm0 xmm2
+
+ /* Restore registers
+ * and exit the function
+ */
L(EXIT):
- addq $72, %rsp
- cfi_def_cfa_offset(8)
- ret
- cfi_def_cfa_offset(80)
+ addq $72, %rsp
+ cfi_def_cfa_offset(8)
+ ret
+ cfi_def_cfa_offset(80)
-/* Branch to process
- * special inputs
- */
+ /* Branch to process
+ * special inputs
+ */
L(SPECIAL_VALUES_BRANCH):
- movups %xmm2, 32(%rsp)
- movups %xmm0, 48(%rsp)
- # LOE rbx rbp r12 r13 r14 r15 edx
-
- xorl %eax, %eax
- movq %r12, 16(%rsp)
- cfi_offset(12, -64)
- movl %eax, %r12d
- movq %r13, 8(%rsp)
- cfi_offset(13, -72)
- movl %edx, %r13d
- movq %r14, (%rsp)
- cfi_offset(14, -80)
- # LOE rbx rbp r15 r12d r13d
-
-/* Range mask
- * bits check
- */
+ movups %xmm2, 32(%rsp)
+ movups %xmm0, 48(%rsp)
+ # LOE rbx rbp r12 r13 r14 r15 edx
+
+ xorl %eax, %eax
+ movq %r12, 16(%rsp)
+ cfi_offset(12, -64)
+ movl %eax, %r12d
+ movq %r13, 8(%rsp)
+ cfi_offset(13, -72)
+ movl %edx, %r13d
+ movq %r14, (%rsp)
+ cfi_offset(14, -80)
+ # LOE rbx rbp 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 rbp r15 r12d r13d
+ /* Call scalar math function */
+ jc L(SCALAR_MATH_CALL)
+ # LOE rbx rbp r15 r12d r13d
-/* Special inputs
- * processing loop
- */
+ /* Special inputs
+ * processing loop
+ */
L(SPECIAL_VALUES_LOOP):
- incl %r12d
- cmpl $4, %r12d
-
-/* Check bits in range mask */
- jl L(RANGEMASK_CHECK)
- # LOE rbx rbp r15 r12d r13d
-
- movq 16(%rsp), %r12
- cfi_restore(12)
- movq 8(%rsp), %r13
- cfi_restore(13)
- movq (%rsp), %r14
- cfi_restore(14)
- movups 48(%rsp), %xmm0
-
-/* Go to exit */
- jmp L(EXIT)
- cfi_offset(12, -64)
- cfi_offset(13, -72)
- cfi_offset(14, -80)
- # LOE rbx rbp r12 r13 r14 r15 xmm0
-
-/* Scalar math fucntion call
- * to process special input
- */
+ incl %r12d
+ cmpl $4, %r12d
+
+ /* Check bits in range mask */
+ jl L(RANGEMASK_CHECK)
+ # LOE rbx rbp r15 r12d r13d
+
+ movq 16(%rsp), %r12
+ cfi_restore(12)
+ movq 8(%rsp), %r13
+ cfi_restore(13)
+ movq (%rsp), %r14
+ cfi_restore(14)
+ movups 48(%rsp), %xmm0
+
+ /* Go to exit */
+ jmp L(EXIT)
+ cfi_offset(12, -64)
+ cfi_offset(13, -72)
+ cfi_offset(14, -80)
+ # LOE rbx rbp r12 r13 r14 r15 xmm0
+
+ /* Scalar math fucntion call
+ * to process special input
+ */
L(SCALAR_MATH_CALL):
- movl %r12d, %r14d
- movss 32(%rsp,%r14,4), %xmm0
- call asinf@PLT
- # LOE rbx rbp r14 r15 r12d r13d xmm0
+ movl %r12d, %r14d
+ movss 32(%rsp, %r14, 4), %xmm0
+ call asinf@PLT
+ # LOE rbx rbp r14 r15 r12d r13d xmm0
- movss %xmm0, 48(%rsp,%r14,4)
+ movss %xmm0, 48(%rsp, %r14, 4)
-/* Process special inputs in loop */
- jmp L(SPECIAL_VALUES_LOOP)
- # LOE rbx rbp r15 r12d r13d
+ /* Process special inputs in loop */
+ jmp L(SPECIAL_VALUES_LOOP)
+ # LOE rbx rbp r15 r12d r13d
END(_ZGVbN4v_asinf_sse4)
- .section .rodata, "a"
- .align 16
+ .section .rodata, "a"
+ .align 16
#ifdef __svml_sasin_data_internal_typedef
typedef unsigned int VUINT32;
typedef struct {
- __declspec(align(16)) VUINT32 AbsMask[4][1];
- __declspec(align(16)) VUINT32 OneHalf[4][1];
- __declspec(align(16)) VUINT32 SmallNorm[4][1];
- __declspec(align(16)) VUINT32 One[4][1];
- __declspec(align(16)) VUINT32 Two[4][1];
- __declspec(align(16)) VUINT32 sqrt_coeff[2][4][1];
- __declspec(align(16)) VUINT32 poly_coeff[5][4][1];
- __declspec(align(16)) VUINT32 Pi2H[4][1];
+ __declspec(align(16)) VUINT32 AbsMask[4][1];
+ __declspec(align(16)) VUINT32 OneHalf[4][1];
+ __declspec(align(16)) VUINT32 SmallNorm[4][1];
+ __declspec(align(16)) VUINT32 One[4][1];
+ __declspec(align(16)) VUINT32 Two[4][1];
+ __declspec(align(16)) VUINT32 sqrt_coeff[2][4][1];
+ __declspec(align(16)) VUINT32 poly_coeff[5][4][1];
+ __declspec(align(16)) VUINT32 Pi2H[4][1];
} __svml_sasin_data_internal;
#endif
__svml_sasin_data_internal:
- /*== AbsMask ==*/
- .long 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff
- /*== OneHalf ==*/
- .align 16
- .long 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000
- /*== SmallNorm ==*/
- .align 16
- .long 0x2f800000, 0x2f800000, 0x2f800000, 0x2f800000
- /*== One ==*/
- .align 16
- .long 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000
- /*== Two ==*/
- .align 16
- .long 0x40000000, 0x40000000, 0x40000000, 0x40000000
- /*== sqrt_coeff[2] ==*/
- .align 16
- .long 0xbdC00004, 0xbdC00004, 0xbdC00004, 0xbdC00004 /* sqrt_coeff2 */
- .long 0x3e800001, 0x3e800001, 0x3e800001, 0x3e800001 /* sqrt_coeff1 */
- /*== poly_coeff[5] ==*/
- .align 16
- .long 0x3d2EDC07, 0x3d2EDC07, 0x3d2EDC07, 0x3d2EDC07 /* poly_coeff5 */
- .long 0x3CC32A6B, 0x3CC32A6B, 0x3CC32A6B, 0x3CC32A6B /* poly_coeff4 */
- .long 0x3d3A9AB4, 0x3d3A9AB4, 0x3d3A9AB4, 0x3d3A9AB4 /* poly_coeff3 */
- .long 0x3d997C12, 0x3d997C12, 0x3d997C12, 0x3d997C12 /* poly_coeff2 */
- .long 0x3e2AAAFF, 0x3e2AAAFF, 0x3e2AAAFF, 0x3e2AAAFF /* poly_coeff1 */
- /*== Pi2H ==*/
- .align 16
- .long 0x3fc90FDB, 0x3fc90FDB, 0x3fc90FDB, 0x3fc90FDB
- .align 16
- .type __svml_sasin_data_internal,@object
- .size __svml_sasin_data_internal,.-__svml_sasin_data_internal
+ /* AbsMask */
+ .long 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff
+ /* OneHalf */
+ .align 16
+ .long 0x3f000000, 0x3f000000, 0x3f000000, 0x3f000000
+ /* SmallNorm */
+ .align 16
+ .long 0x2f800000, 0x2f800000, 0x2f800000, 0x2f800000
+ /* One */
+ .align 16
+ .long 0x3f800000, 0x3f800000, 0x3f800000, 0x3f800000
+ /* Two */
+ .align 16
+ .long 0x40000000, 0x40000000, 0x40000000, 0x40000000
+ /* sqrt_coeff[2] */
+ .align 16
+ .long 0xbdC00004, 0xbdC00004, 0xbdC00004, 0xbdC00004 /* sqrt_coeff2 */
+ .long 0x3e800001, 0x3e800001, 0x3e800001, 0x3e800001 /* sqrt_coeff1 */
+ /* poly_coeff[5] */
+ .align 16
+ .long 0x3d2EDC07, 0x3d2EDC07, 0x3d2EDC07, 0x3d2EDC07 /* poly_coeff5 */
+ .long 0x3CC32A6B, 0x3CC32A6B, 0x3CC32A6B, 0x3CC32A6B /* poly_coeff4 */
+ .long 0x3d3A9AB4, 0x3d3A9AB4, 0x3d3A9AB4, 0x3d3A9AB4 /* poly_coeff3 */
+ .long 0x3d997C12, 0x3d997C12, 0x3d997C12, 0x3d997C12 /* poly_coeff2 */
+ .long 0x3e2AAAFF, 0x3e2AAAFF, 0x3e2AAAFF, 0x3e2AAAFF /* poly_coeff1 */
+ /* Pi2H */
+ .align 16
+ .long 0x3fc90FDB, 0x3fc90FDB, 0x3fc90FDB, 0x3fc90FDB
+ .align 16
+ .type __svml_sasin_data_internal, @object
+ .size __svml_sasin_data_internal, .-__svml_sasin_data_internal