mirror of git://sourceware.org/git/glibc.git
x86_64: Fix svml_s_exp2f16_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 <goldstein.w.n@gmail.com>
This commit is contained in:
Sunil K Pandey
parent
7998aecb4d
commit
eb927a26fa
|
|
@ -62,8 +62,7 @@
|
|||
|
||||
#include <sysdep.h>
|
||||
|
||||
.text
|
||||
.section .text.exex512,"ax",@progbits
|
||||
.section .text.exex512, "ax", @progbits
|
||||
ENTRY(_ZGVeN16v_exp2f_skx)
|
||||
pushq %rbp
|
||||
cfi_def_cfa_offset(16)
|
||||
|
|
@ -74,7 +73,7 @@ ENTRY(_ZGVeN16v_exp2f_skx)
|
|||
subq $192, %rsp
|
||||
vmovups add_const+__svml_sexp2_data_internal_avx512(%rip), %zmm3
|
||||
|
||||
/*
|
||||
/*
|
||||
* Reduced argument
|
||||
* where VREDUCE is available
|
||||
*/
|
||||
|
|
@ -83,7 +82,7 @@ ENTRY(_ZGVeN16v_exp2f_skx)
|
|||
vmovups poly_coeff2+__svml_sexp2_data_internal_avx512(%rip), %zmm10
|
||||
vmovups Threshold+__svml_sexp2_data_internal_avx512(%rip), %zmm2
|
||||
|
||||
/*
|
||||
/*
|
||||
*
|
||||
* HA
|
||||
* Variables and constants
|
||||
|
|
@ -91,39 +90,39 @@ ENTRY(_ZGVeN16v_exp2f_skx)
|
|||
*/
|
||||
vmovups poly_coeff1+__svml_sexp2_data_internal_avx512(%rip), %zmm7
|
||||
|
||||
/*
|
||||
/*
|
||||
* Integer form of K+0.b1b2b3b4 in lower bits - call K_plus_f0
|
||||
* Mantisssa of normalized single precision FP: 1.b1b2...b23
|
||||
*/
|
||||
vaddps {rd-sae}, %zmm3, %zmm0, %zmm4
|
||||
vandps AbsMask+__svml_sexp2_data_internal_avx512(%rip), %zmm0, %zmm1
|
||||
|
||||
/* c3*r + c2 */
|
||||
/* c3*r + c2 */
|
||||
vfmadd231ps {rn-sae}, %zmm6, %zmm5, %zmm10
|
||||
vcmpps $30, {sae}, %zmm2, %zmm1, %k0
|
||||
|
||||
/* c3*r^2 + c2*r + c1 */
|
||||
/* c3*r^2 + c2*r + c1 */
|
||||
vfmadd213ps {rn-sae}, %zmm7, %zmm6, %zmm10
|
||||
|
||||
/* Table value: 2^(0.b1b2b3b4) */
|
||||
/* Table value: 2^(0.b1b2b3b4) */
|
||||
vpermps __svml_sexp2_data_internal_avx512(%rip), %zmm4, %zmm9
|
||||
kmovw %k0, %edx
|
||||
|
||||
/* T*r */
|
||||
/* T*r */
|
||||
vmulps {rn-sae}, %zmm6, %zmm9, %zmm8
|
||||
|
||||
/* T + (T*r*(c3*r^2 + c2*r + c1) */
|
||||
/* T + (T*r*(c3*r^2 + c2*r + c1) */
|
||||
vfmadd213ps {rn-sae}, %zmm9, %zmm8, %zmm10
|
||||
|
||||
/* Scaling placed at the end to avoid accuracy loss when T*r*scale underflows */
|
||||
/* Scaling placed at the end to avoid accuracy loss when T*r*scale underflows */
|
||||
vscalefps {rn-sae}, %zmm0, %zmm10, %zmm1
|
||||
testl %edx, %edx
|
||||
|
||||
/* Go to special inputs processing branch */
|
||||
/* Go to special inputs processing branch */
|
||||
jne L(SPECIAL_VALUES_BRANCH)
|
||||
# LOE rbx r12 r13 r14 r15 edx zmm0 zmm1
|
||||
|
||||
/* Restore registers
|
||||
/* Restore registers
|
||||
* and exit the function
|
||||
*/
|
||||
|
||||
|
|
@ -137,7 +136,7 @@ L(EXIT):
|
|||
cfi_def_cfa(6, 16)
|
||||
cfi_offset(6, -16)
|
||||
|
||||
/* Branch to process
|
||||
/* Branch to process
|
||||
* special inputs
|
||||
*/
|
||||
|
||||
|
|
@ -163,18 +162,18 @@ L(SPECIAL_VALUES_BRANCH):
|
|||
.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
|
||||
/* Range mask
|
||||
* bits check
|
||||
*/
|
||||
|
||||
L(RANGEMASK_CHECK):
|
||||
btl %r12d, %r13d
|
||||
|
||||
/* Call scalar math function */
|
||||
/* Call scalar math function */
|
||||
jc L(SCALAR_MATH_CALL)
|
||||
# LOE rbx r15 r12d r13d
|
||||
|
||||
/* Special inputs
|
||||
/* Special inputs
|
||||
* processing loop
|
||||
*/
|
||||
|
||||
|
|
@ -182,7 +181,7 @@ L(SPECIAL_VALUES_LOOP):
|
|||
incl %r12d
|
||||
cmpl $16, %r12d
|
||||
|
||||
/* Check bits in range mask */
|
||||
/* Check bits in range mask */
|
||||
jl L(RANGEMASK_CHECK)
|
||||
# LOE rbx r15 r12d r13d
|
||||
|
||||
|
|
@ -194,7 +193,7 @@ L(SPECIAL_VALUES_LOOP):
|
|||
cfi_restore(14)
|
||||
vmovups 128(%rsp), %zmm1
|
||||
|
||||
/* Go to exit */
|
||||
/* 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
|
||||
|
|
@ -204,19 +203,19 @@ L(SPECIAL_VALUES_LOOP):
|
|||
.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 zmm1
|
||||
|
||||
/* Scalar math fucntion call
|
||||
/* Scalar math fucntion call
|
||||
* to process special input
|
||||
*/
|
||||
|
||||
L(SCALAR_MATH_CALL):
|
||||
movl %r12d, %r14d
|
||||
movss 64(%rsp,%r14,4), %xmm0
|
||||
movss 64(%rsp, %r14, 4), %xmm0
|
||||
call exp2f@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 */
|
||||
/* Process special inputs in loop */
|
||||
jmp L(SPECIAL_VALUES_LOOP)
|
||||
# LOE rbx r15 r12d r13d
|
||||
END(_ZGVeN16v_exp2f_skx)
|
||||
|
|
@ -237,7 +236,7 @@ typedef struct {
|
|||
} __svml_sexp2_data_internal_avx512;
|
||||
#endif
|
||||
__svml_sexp2_data_internal_avx512:
|
||||
/*== Frac_PowerS0 ==*/
|
||||
/* Frac_PowerS0 */
|
||||
.long 0x3F800000
|
||||
.long 0x3F85AAC3
|
||||
.long 0x3F8B95C2
|
||||
|
|
@ -255,11 +254,11 @@ __svml_sexp2_data_internal_avx512:
|
|||
.long 0x3FEAC0C7
|
||||
.long 0x3FF5257D
|
||||
.align 64
|
||||
.long 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222 /*== poly_coeff1 ==*/
|
||||
.long 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222, 0x3F317222 /* == poly_coeff1 == */
|
||||
.align 64
|
||||
.long 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B /*== poly_coeff2 ==*/
|
||||
.long 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B, 0x3E75F16B /* == poly_coeff2 == */
|
||||
.align 64
|
||||
.long 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA /*== poly_coeff3 ==*/
|
||||
.long 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA, 0x3D6854CA /* == poly_coeff3 == */
|
||||
.align 64
|
||||
.long 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000, 0x49400000 /* add_const */
|
||||
.align 64
|
||||
|
|
@ -267,5 +266,5 @@ __svml_sexp2_data_internal_avx512:
|
|||
.align 64
|
||||
.long 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000, 0x42fc0000 /* Threshold=126.0 */
|
||||
.align 64
|
||||
.type __svml_sexp2_data_internal_avx512,@object
|
||||
.size __svml_sexp2_data_internal_avx512,.-__svml_sexp2_data_internal_avx512
|
||||
.type __svml_sexp2_data_internal_avx512, @object
|
||||
.size __svml_sexp2_data_internal_avx512, .-__svml_sexp2_data_internal_avx512
|
||||
|
|
|
|||
Loading…
Reference in New Issue