aboutsummaryrefslogtreecommitdiff
path: root/sysdeps
diff options
context:
space:
mode:
authorSunil K Pandey <skpgkp2@gmail.com>2022-03-07 10:47:09 -0800
committerSunil K Pandey <skpgkp2@gmail.com>2022-03-07 21:14:09 -0800
commitb1e3e51b0b0f4f0c2c11d473ee8d4b278779c590 (patch)
tree2a1d0729714d2ab34000777fe774388cbc779643 /sysdeps
parent14dd6fa8b4d320b1991e5ed2ad3ad060d669e982 (diff)
downloadglibc-b1e3e51b0b0f4f0c2c11d473ee8d4b278779c590.tar
glibc-b1e3e51b0b0f4f0c2c11d473ee8d4b278779c590.tar.gz
glibc-b1e3e51b0b0f4f0c2c11d473ee8d4b278779c590.tar.bz2
glibc-b1e3e51b0b0f4f0c2c11d473ee8d4b278779c590.zip
x86_64: Fix svml_d_atan2_core_sse4.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 <goldstein.w.n@gmail.com>
Diffstat (limited to 'sysdeps')
-rw-r--r--sysdeps/x86_64/fpu/multiarch/svml_d_atan2_core_sse4.S395
1 files changed, 197 insertions, 198 deletions
diff --git a/sysdeps/x86_64/fpu/multiarch/svml_d_atan2_core_sse4.S b/sysdeps/x86_64/fpu/multiarch/svml_d_atan2_core_sse4.S
index 31194c8e4c..055ca4acc8 100644
--- a/sysdeps/x86_64/fpu/multiarch/svml_d_atan2_core_sse4.S
+++ b/sysdeps/x86_64/fpu/multiarch/svml_d_atan2_core_sse4.S
@@ -30,216 +30,215 @@
/* Offsets for data table __svml_datan_data_internal_avx512
*/
-#define AbsMask 0
-#define Shifter 16
-#define MaxThreshold 32
-#define MOne 48
-#define One 64
-#define LargeX 80
-#define Zero 96
-#define Tbl_H 112
-#define Tbl_L 368
-#define dIndexMed 624
-#define Pi2 640
-#define Pi2_low 656
-#define coeff 672
+#define AbsMask 0
+#define Shifter 16
+#define MaxThreshold 32
+#define MOne 48
+#define One 64
+#define LargeX 80
+#define Zero 96
+#define Tbl_H 112
+#define Tbl_L 368
+#define dIndexMed 624
+#define Pi2 640
+#define Pi2_low 656
+#define coeff 672
#include <sysdep.h>
- .text
- .section .text.sse4,"ax",@progbits
+ .section .text.sse4, "ax", @progbits
ENTRY(_ZGVbN2v_atan_sse4)
- lea Tbl_H+128+__svml_datan_data_internal_avx512(%rip), %rcx
- movups __svml_datan_data_internal_avx512(%rip), %xmm4
- movups Shifter+__svml_datan_data_internal_avx512(%rip), %xmm3
- andps %xmm0, %xmm4
- movaps %xmm3, %xmm12
- movaps %xmm4, %xmm5
- addpd %xmm4, %xmm12
- movaps %xmm12, %xmm7
-
-/*
- * table lookup sequence
- * VPERMUTE not available
- */
- movaps %xmm12, %xmm10
- subpd %xmm3, %xmm7
- subpd %xmm7, %xmm5
- mulpd %xmm4, %xmm7
- movups MaxThreshold+__svml_datan_data_internal_avx512(%rip), %xmm2
- psllq $3, %xmm10
-
-/* saturate X range */
- movups LargeX+__svml_datan_data_internal_avx512(%rip), %xmm8
- pxor %xmm4, %xmm0
- cmplepd %xmm4, %xmm2
- addpd One+__svml_datan_data_internal_avx512(%rip), %xmm7
- minpd %xmm4, %xmm8
- movups MOne+__svml_datan_data_internal_avx512(%rip), %xmm6
- movaps %xmm2, %xmm1
- movaps %xmm2, %xmm9
- andnps %xmm5, %xmm1
- andps %xmm2, %xmm6
- andnps %xmm7, %xmm9
- andps %xmm2, %xmm8
- orps %xmm6, %xmm1
- orps %xmm8, %xmm9
-
-/* R+Rl = DiffX/Y */
- divpd %xmm9, %xmm1
- pand .FLT_11(%rip), %xmm10
-
-/* set table value to Pi/2 for large X */
- movups Pi2+__svml_datan_data_internal_avx512(%rip), %xmm4
- movd %xmm10, %eax
- andps %xmm2, %xmm4
- pshufd $2, %xmm10, %xmm11
- movaps %xmm2, %xmm10
-
-/* polynomial evaluation */
- movaps %xmm1, %xmm2
- mulpd %xmm1, %xmm2
- movd %xmm11, %edx
- movups coeff+__svml_datan_data_internal_avx512(%rip), %xmm5
- movaps %xmm2, %xmm7
- movups coeff+32+__svml_datan_data_internal_avx512(%rip), %xmm6
- movaps %xmm2, %xmm9
- mulpd %xmm2, %xmm5
- mulpd %xmm2, %xmm7
- addpd coeff+16+__svml_datan_data_internal_avx512(%rip), %xmm5
- mulpd %xmm2, %xmm6
- mulpd %xmm7, %xmm5
- addpd coeff+48+__svml_datan_data_internal_avx512(%rip), %xmm6
- mulpd %xmm1, %xmm9
- addpd %xmm5, %xmm6
- movups coeff+64+__svml_datan_data_internal_avx512(%rip), %xmm8
- mulpd %xmm2, %xmm8
- mulpd %xmm6, %xmm7
- addpd coeff+80+__svml_datan_data_internal_avx512(%rip), %xmm8
- addpd %xmm7, %xmm8
- mulpd %xmm8, %xmm9
- movups dIndexMed+__svml_datan_data_internal_avx512(%rip), %xmm14
- cmplepd %xmm12, %xmm14
- addpd %xmm9, %xmm1
- movslq %eax, %rax
- movaps %xmm14, %xmm3
- movslq %edx, %rdx
- movsd -128(%rax,%rcx), %xmm13
- movsd (%rcx,%rax), %xmm15
- movhpd -128(%rdx,%rcx), %xmm13
- movhpd (%rcx,%rdx), %xmm15
- andnps %xmm13, %xmm3
- andps %xmm14, %xmm15
- orps %xmm15, %xmm3
- andnps %xmm3, %xmm10
- orps %xmm4, %xmm10
- addpd %xmm1, %xmm10
- pxor %xmm10, %xmm0
- ret
+ lea Tbl_H+128+__svml_datan_data_internal_avx512(%rip), %rcx
+ movups __svml_datan_data_internal_avx512(%rip), %xmm4
+ movups Shifter+__svml_datan_data_internal_avx512(%rip), %xmm3
+ andps %xmm0, %xmm4
+ movaps %xmm3, %xmm12
+ movaps %xmm4, %xmm5
+ addpd %xmm4, %xmm12
+ movaps %xmm12, %xmm7
+
+ /*
+ * table lookup sequence
+ * VPERMUTE not available
+ */
+ movaps %xmm12, %xmm10
+ subpd %xmm3, %xmm7
+ subpd %xmm7, %xmm5
+ mulpd %xmm4, %xmm7
+ movups MaxThreshold+__svml_datan_data_internal_avx512(%rip), %xmm2
+ psllq $3, %xmm10
+
+ /* saturate X range */
+ movups LargeX+__svml_datan_data_internal_avx512(%rip), %xmm8
+ pxor %xmm4, %xmm0
+ cmplepd %xmm4, %xmm2
+ addpd One+__svml_datan_data_internal_avx512(%rip), %xmm7
+ minpd %xmm4, %xmm8
+ movups MOne+__svml_datan_data_internal_avx512(%rip), %xmm6
+ movaps %xmm2, %xmm1
+ movaps %xmm2, %xmm9
+ andnps %xmm5, %xmm1
+ andps %xmm2, %xmm6
+ andnps %xmm7, %xmm9
+ andps %xmm2, %xmm8
+ orps %xmm6, %xmm1
+ orps %xmm8, %xmm9
+
+ /* R+Rl = DiffX/Y */
+ divpd %xmm9, %xmm1
+ pand .FLT_11(%rip), %xmm10
+
+ /* set table value to Pi/2 for large X */
+ movups Pi2+__svml_datan_data_internal_avx512(%rip), %xmm4
+ movd %xmm10, %eax
+ andps %xmm2, %xmm4
+ pshufd $2, %xmm10, %xmm11
+ movaps %xmm2, %xmm10
+
+ /* polynomial evaluation */
+ movaps %xmm1, %xmm2
+ mulpd %xmm1, %xmm2
+ movd %xmm11, %edx
+ movups coeff+__svml_datan_data_internal_avx512(%rip), %xmm5
+ movaps %xmm2, %xmm7
+ movups coeff+32+__svml_datan_data_internal_avx512(%rip), %xmm6
+ movaps %xmm2, %xmm9
+ mulpd %xmm2, %xmm5
+ mulpd %xmm2, %xmm7
+ addpd coeff+16+__svml_datan_data_internal_avx512(%rip), %xmm5
+ mulpd %xmm2, %xmm6
+ mulpd %xmm7, %xmm5
+ addpd coeff+48+__svml_datan_data_internal_avx512(%rip), %xmm6
+ mulpd %xmm1, %xmm9
+ addpd %xmm5, %xmm6
+ movups coeff+64+__svml_datan_data_internal_avx512(%rip), %xmm8
+ mulpd %xmm2, %xmm8
+ mulpd %xmm6, %xmm7
+ addpd coeff+80+__svml_datan_data_internal_avx512(%rip), %xmm8
+ addpd %xmm7, %xmm8
+ mulpd %xmm8, %xmm9
+ movups dIndexMed+__svml_datan_data_internal_avx512(%rip), %xmm14
+ cmplepd %xmm12, %xmm14
+ addpd %xmm9, %xmm1
+ movslq %eax, %rax
+ movaps %xmm14, %xmm3
+ movslq %edx, %rdx
+ movsd -128(%rax, %rcx), %xmm13
+ movsd (%rcx, %rax), %xmm15
+ movhpd -128(%rdx, %rcx), %xmm13
+ movhpd (%rcx, %rdx), %xmm15
+ andnps %xmm13, %xmm3
+ andps %xmm14, %xmm15
+ orps %xmm15, %xmm3
+ andnps %xmm3, %xmm10
+ orps %xmm4, %xmm10
+ addpd %xmm1, %xmm10
+ pxor %xmm10, %xmm0
+ ret
END(_ZGVbN2v_atan_sse4)
- .section .rodata, "a"
- .align 16
+ .section .rodata, "a"
+ .align 16
#ifdef __svml_datan_data_internal_avx512_typedef
typedef unsigned int VUINT32;
typedef struct {
- __declspec(align(16)) VUINT32 AbsMask[2][2];
- __declspec(align(16)) VUINT32 Shifter[2][2];
- __declspec(align(16)) VUINT32 MaxThreshold[2][2];
- __declspec(align(16)) VUINT32 MOne[2][2];
- __declspec(align(16)) VUINT32 One[2][2];
- __declspec(align(16)) VUINT32 LargeX[2][2];
- __declspec(align(16)) VUINT32 Zero[2][2];
- __declspec(align(16)) VUINT32 Tbl_H[32][2];
- __declspec(align(16)) VUINT32 Tbl_L[32][2];
- __declspec(align(16)) VUINT32 dIndexMed[2][2];
- __declspec(align(16)) VUINT32 Pi2[2][2];
- __declspec(align(16)) VUINT32 Pi2_low[2][2];
- __declspec(align(16)) VUINT32 coeff[6][2][2];
- } __svml_datan_data_internal_avx512;
+ __declspec(align(16)) VUINT32 AbsMask[2][2];
+ __declspec(align(16)) VUINT32 Shifter[2][2];
+ __declspec(align(16)) VUINT32 MaxThreshold[2][2];
+ __declspec(align(16)) VUINT32 MOne[2][2];
+ __declspec(align(16)) VUINT32 One[2][2];
+ __declspec(align(16)) VUINT32 LargeX[2][2];
+ __declspec(align(16)) VUINT32 Zero[2][2];
+ __declspec(align(16)) VUINT32 Tbl_H[32][2];
+ __declspec(align(16)) VUINT32 Tbl_L[32][2];
+ __declspec(align(16)) VUINT32 dIndexMed[2][2];
+ __declspec(align(16)) VUINT32 Pi2[2][2];
+ __declspec(align(16)) VUINT32 Pi2_low[2][2];
+ __declspec(align(16)) VUINT32 coeff[6][2][2];
+} __svml_datan_data_internal_avx512;
#endif
__svml_datan_data_internal_avx512:
- /*== AbsMask ==*/
- .quad 0x7fffffffffffffff, 0x7fffffffffffffff
- /*== Shifter ==*/
- .align 16
- .quad 0x4318000000000000, 0x4318000000000000
- /*== MaxThreshold ==*/
- .align 16
- .quad 0x401f800000000000, 0x401f800000000000
- /*== MOne ==*/
- .align 16
- .quad 0xbff0000000000000, 0xbff0000000000000
- /*== One ==*/
- .align 16
- .quad 0x3ff0000000000000, 0x3ff0000000000000
- /*== LargeX ==*/
- .align 16
- .quad 0x47f0000000000000, 0x47f0000000000000
- /*== Zero ==*/
- .align 16
- .quad 0x0000000000000000, 0x0000000000000000
- /*== Tbl_H ==*/
- .align 16
- .quad 0x0000000000000000, 0x3fcf5b75f92c80dd
- .quad 0x3fddac670561bb4f, 0x3fe4978fa3269ee1
- .quad 0x3fe921fb54442d18, 0x3fecac7c57846f9e
- .quad 0x3fef730bd281f69b, 0x3ff0d38f2c5ba09f
- .quad 0x3ff1b6e192ebbe44, 0x3ff270ef55a53a25
- .quad 0x3ff30b6d796a4da8, 0x3ff38d6a6ce13353
- .quad 0x3ff3fc176b7a8560, 0x3ff45b54837351a0
- .quad 0x3ff4ae10fc6589a5, 0x3ff4f68dea672617
- .quad 0x3ff5368c951e9cfd, 0x3ff56f6f33a3e6a7
- .quad 0x3ff5a25052114e60, 0x3ff5d013c41adabd
- .quad 0x3ff5f97315254857, 0x3ff61f06c6a92b89
- .quad 0x3ff6414d44094c7c, 0x3ff660b02c736a06
- .quad 0x3ff67d8863bc99bd, 0x3ff698213a9d5053
- .quad 0x3ff6b0bae830c070, 0x3ff6c78c7edeb195
- .quad 0x3ff6dcc57bb565fd, 0x3ff6f08f07435fec
- .quad 0x3ff7030cf9403197, 0x3ff7145eac2088a4
- /*== Tbl_L ==*/
- .align 16
- .quad 0x0000000000000000, 0x3c68ab6e3cf7afbd
- .quad 0x3c7a2b7f222f65e2, 0x3c72419a87f2a458
- .quad 0x3c81a62633145c07, 0x3c80dae13ad18a6b
- .quad 0x3c7007887af0cbbd, 0xbc9bd0dc231bfd70
- .quad 0x3c9b1b466a88828e, 0xbc9a66b1af5f84fb
- .quad 0x3c96254cb03bb199, 0xbc812c77e8a80f5c
- .quad 0xbc4441a3bd3f1084, 0x3c79e4a72eedacc4
- .quad 0xbc93b03e8a27f555, 0x3c9934f9f2b0020e
- .quad 0xbc996f47948a99f1, 0xbc7df6edd6f1ec3b
- .quad 0x3c78c2d0c89de218, 0x3c9f82bba194dd5d
- .quad 0xbc831151a43b51ca, 0xbc8487d50bceb1a5
- .quad 0xbc9c5f60a65c7397, 0xbc7acb6afb332a0f
- .quad 0xbc99b7bd2e1e8c9c, 0xbc9b9839085189e3
- .quad 0xbc97d1ab82ffb70b, 0x3c99239ad620ffe2
- .quad 0xbc929c86447928e7, 0xbc8957a7170df016
- .quad 0xbc7cbe1896221608, 0xbc9fda5797b32a0b
- /*== dIndexMed ==*/
- .align 16
- .quad 0x4318000000000010, 0x4318000000000010
- /*== Pi2 ==*/
- .align 16
- .quad 0x3ff921fb54442d18, 0x3ff921fb54442d18
- /*== Pi2_low ==*/
- .align 16
- .quad 0x3c91a62633145c07, 0x3c91a62633145c07
- /*== coeff6 ==*/
- .align 16
- .quad 0x3fb2e9b9f5c4fe97, 0x3fb2e9b9f5c4fe97
- .quad 0xbfb74257c46790cc, 0xbfb74257c46790cc
- .quad 0x3fbc71bfeff916a0, 0x3fbc71bfeff916a0
- .quad 0xbfc249248eef04da, 0xbfc249248eef04da
- .quad 0x3fc999999998741e, 0x3fc999999998741e
- .quad 0xbfd555555555554d, 0xbfd555555555554d
- .align 16
- .type __svml_datan_data_internal_avx512,@object
- .size __svml_datan_data_internal_avx512,.-__svml_datan_data_internal_avx512
- .align 16
+ /* AbsMask */
+ .quad 0x7fffffffffffffff, 0x7fffffffffffffff
+ /* Shifter */
+ .align 16
+ .quad 0x4318000000000000, 0x4318000000000000
+ /* MaxThreshold */
+ .align 16
+ .quad 0x401f800000000000, 0x401f800000000000
+ /* MOne */
+ .align 16
+ .quad 0xbff0000000000000, 0xbff0000000000000
+ /* One */
+ .align 16
+ .quad 0x3ff0000000000000, 0x3ff0000000000000
+ /* LargeX */
+ .align 16
+ .quad 0x47f0000000000000, 0x47f0000000000000
+ /* Zero */
+ .align 16
+ .quad 0x0000000000000000, 0x0000000000000000
+ /* Tbl_H */
+ .align 16
+ .quad 0x0000000000000000, 0x3fcf5b75f92c80dd
+ .quad 0x3fddac670561bb4f, 0x3fe4978fa3269ee1
+ .quad 0x3fe921fb54442d18, 0x3fecac7c57846f9e
+ .quad 0x3fef730bd281f69b, 0x3ff0d38f2c5ba09f
+ .quad 0x3ff1b6e192ebbe44, 0x3ff270ef55a53a25
+ .quad 0x3ff30b6d796a4da8, 0x3ff38d6a6ce13353
+ .quad 0x3ff3fc176b7a8560, 0x3ff45b54837351a0
+ .quad 0x3ff4ae10fc6589a5, 0x3ff4f68dea672617
+ .quad 0x3ff5368c951e9cfd, 0x3ff56f6f33a3e6a7
+ .quad 0x3ff5a25052114e60, 0x3ff5d013c41adabd
+ .quad 0x3ff5f97315254857, 0x3ff61f06c6a92b89
+ .quad 0x3ff6414d44094c7c, 0x3ff660b02c736a06
+ .quad 0x3ff67d8863bc99bd, 0x3ff698213a9d5053
+ .quad 0x3ff6b0bae830c070, 0x3ff6c78c7edeb195
+ .quad 0x3ff6dcc57bb565fd, 0x3ff6f08f07435fec
+ .quad 0x3ff7030cf9403197, 0x3ff7145eac2088a4
+ /* Tbl_L */
+ .align 16
+ .quad 0x0000000000000000, 0x3c68ab6e3cf7afbd
+ .quad 0x3c7a2b7f222f65e2, 0x3c72419a87f2a458
+ .quad 0x3c81a62633145c07, 0x3c80dae13ad18a6b
+ .quad 0x3c7007887af0cbbd, 0xbc9bd0dc231bfd70
+ .quad 0x3c9b1b466a88828e, 0xbc9a66b1af5f84fb
+ .quad 0x3c96254cb03bb199, 0xbc812c77e8a80f5c
+ .quad 0xbc4441a3bd3f1084, 0x3c79e4a72eedacc4
+ .quad 0xbc93b03e8a27f555, 0x3c9934f9f2b0020e
+ .quad 0xbc996f47948a99f1, 0xbc7df6edd6f1ec3b
+ .quad 0x3c78c2d0c89de218, 0x3c9f82bba194dd5d
+ .quad 0xbc831151a43b51ca, 0xbc8487d50bceb1a5
+ .quad 0xbc9c5f60a65c7397, 0xbc7acb6afb332a0f
+ .quad 0xbc99b7bd2e1e8c9c, 0xbc9b9839085189e3
+ .quad 0xbc97d1ab82ffb70b, 0x3c99239ad620ffe2
+ .quad 0xbc929c86447928e7, 0xbc8957a7170df016
+ .quad 0xbc7cbe1896221608, 0xbc9fda5797b32a0b
+ /* dIndexMed */
+ .align 16
+ .quad 0x4318000000000010, 0x4318000000000010
+ /* Pi2 */
+ .align 16
+ .quad 0x3ff921fb54442d18, 0x3ff921fb54442d18
+ /* Pi2_low */
+ .align 16
+ .quad 0x3c91a62633145c07, 0x3c91a62633145c07
+ /* coeff6 */
+ .align 16
+ .quad 0x3fb2e9b9f5c4fe97, 0x3fb2e9b9f5c4fe97
+ .quad 0xbfb74257c46790cc, 0xbfb74257c46790cc
+ .quad 0x3fbc71bfeff916a0, 0x3fbc71bfeff916a0
+ .quad 0xbfc249248eef04da, 0xbfc249248eef04da
+ .quad 0x3fc999999998741e, 0x3fc999999998741e
+ .quad 0xbfd555555555554d, 0xbfd555555555554d
+ .align 16
+ .type __svml_datan_data_internal_avx512, @object
+ .size __svml_datan_data_internal_avx512, .-__svml_datan_data_internal_avx512
+ .align 16
.FLT_11:
- .long 0x00000078,0x00000000,0x00000078,0x00000000
- .type .FLT_11,@object
- .size .FLT_11,16
+ .long 0x00000078, 0x00000000, 0x00000078, 0x00000000
+ .type .FLT_11, @object
+ .size .FLT_11, 16