mirror of
git://sourceware.org/git/glibc.git
synced 2024-11-21 01:12:26 +08:00
x86_64: Fix svml_s_exp10f4_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>
This commit is contained in:
Sunil K Pandey
parent
00e651209b
commit
dcc7a3e02d
@ -22,7 +22,7 @@
|
||||
* exp10(x) = 2^x/log10(2) = 2^n * (1 + T[j]) * (1 + P(y))
|
||||
* where
|
||||
* x = m*log10(2)/K + y, y in [-log10(2)/K..log10(2)/K]
|
||||
* m = n*K + j, m,n,j - signed integer, j in [-K/2..K/2]
|
||||
* m = n*K + j, m, n,j - signed integer, j in [-K/2..K/2]
|
||||
*
|
||||
* values of 2^j/K are tabulated
|
||||
*
|
||||
@ -57,20 +57,19 @@
|
||||
|
||||
#include <sysdep.h>
|
||||
|
||||
.text
|
||||
.section .text.sse4,"ax",@progbits
|
||||
.section .text.sse4, "ax", @progbits
|
||||
ENTRY(_ZGVbN4v_exp10f_sse4)
|
||||
subq $72, %rsp
|
||||
cfi_def_cfa_offset(80)
|
||||
movaps %xmm0, %xmm4
|
||||
|
||||
/* Load arument */
|
||||
/* Load arument */
|
||||
movups _sLg2_10+__svml_sexp10_data_internal(%rip), %xmm2
|
||||
lea __svml_sexp10_data_internal(%rip), %r8
|
||||
mulps %xmm4, %xmm2
|
||||
movups _sShifter+__svml_sexp10_data_internal(%rip), %xmm5
|
||||
|
||||
/* R */
|
||||
/* R */
|
||||
movups _sInvLg2_10hi+__svml_sexp10_data_internal(%rip), %xmm14
|
||||
addps %xmm5, %xmm2
|
||||
movaps %xmm2, %xmm1
|
||||
@ -81,7 +80,7 @@ ENTRY(_ZGVbN4v_exp10f_sse4)
|
||||
mulps %xmm1, %xmm15
|
||||
subps %xmm14, %xmm5
|
||||
|
||||
/*
|
||||
/*
|
||||
* Polynomial
|
||||
* exp10 = 2^N*(Tj+Tj*poly)
|
||||
* poly(sN) = {1+later} a0+a1*sR
|
||||
@ -91,17 +90,17 @@ ENTRY(_ZGVbN4v_exp10f_sse4)
|
||||
mulps %xmm5, %xmm1
|
||||
movdqu _iIndexMask+__svml_sexp10_data_internal(%rip), %xmm3
|
||||
|
||||
/* Index and lookup */
|
||||
/* Index and lookup */
|
||||
movdqa %xmm3, %xmm10
|
||||
|
||||
/* remove index bits */
|
||||
/* remove index bits */
|
||||
pandn %xmm2, %xmm3
|
||||
pand %xmm2, %xmm10
|
||||
|
||||
/* 2^N */
|
||||
/* 2^N */
|
||||
pslld $18, %xmm3
|
||||
|
||||
/* iIndex *= sizeof(S); */
|
||||
/* iIndex *= sizeof(S); */
|
||||
pslld $2, %xmm10
|
||||
addps _sPC1+__svml_sexp10_data_internal(%rip), %xmm1
|
||||
movd %xmm10, %edx
|
||||
@ -112,7 +111,7 @@ ENTRY(_ZGVbN4v_exp10f_sse4)
|
||||
movd %xmm9, %esi
|
||||
movd %xmm11, %edi
|
||||
|
||||
/* Check for overflow\underflow */
|
||||
/* Check for overflow\underflow */
|
||||
movdqu _iAbsMask+__svml_sexp10_data_internal(%rip), %xmm6
|
||||
pand %xmm4, %xmm6
|
||||
mulps %xmm1, %xmm5
|
||||
@ -121,31 +120,31 @@ ENTRY(_ZGVbN4v_exp10f_sse4)
|
||||
movslq %ecx, %rcx
|
||||
movslq %esi, %rsi
|
||||
movslq %edi, %rdi
|
||||
movd (%r8,%rdx), %xmm0
|
||||
movd (%r8,%rcx), %xmm8
|
||||
movd (%r8,%rsi), %xmm13
|
||||
movd (%r8,%rdi), %xmm12
|
||||
movd (%r8, %rdx), %xmm0
|
||||
movd (%r8, %rcx), %xmm8
|
||||
movd (%r8, %rsi), %xmm13
|
||||
movd (%r8, %rdi), %xmm12
|
||||
punpckldq %xmm8, %xmm0
|
||||
punpckldq %xmm12, %xmm13
|
||||
punpcklqdq %xmm13, %xmm0
|
||||
|
||||
/* Tj_l+Tj_h*poly */
|
||||
/* Tj_l+Tj_h*poly */
|
||||
mulps %xmm0, %xmm5
|
||||
pcmpgtd _iDomainRange+__svml_sexp10_data_internal(%rip), %xmm6
|
||||
addps %xmm5, %xmm0
|
||||
movmskps %xmm6, %eax
|
||||
|
||||
/* quick mul 2^N */
|
||||
/* quick mul 2^N */
|
||||
paddd %xmm3, %xmm0
|
||||
|
||||
/* Finish */
|
||||
/* Finish */
|
||||
testl %eax, %eax
|
||||
|
||||
/* Go to special inputs processing branch */
|
||||
/* Go to special inputs processing branch */
|
||||
jne L(SPECIAL_VALUES_BRANCH)
|
||||
# LOE rbx rbp r12 r13 r14 r15 eax xmm0 xmm4
|
||||
|
||||
/* Restore registers
|
||||
/* Restore registers
|
||||
* and exit the function
|
||||
*/
|
||||
|
||||
@ -155,7 +154,7 @@ L(EXIT):
|
||||
ret
|
||||
cfi_def_cfa_offset(80)
|
||||
|
||||
/* Branch to process
|
||||
/* Branch to process
|
||||
* special inputs
|
||||
*/
|
||||
|
||||
@ -175,18 +174,18 @@ L(SPECIAL_VALUES_BRANCH):
|
||||
cfi_offset(14, -80)
|
||||
# LOE rbx rbp 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 rbp r15 r12d r13d
|
||||
|
||||
/* Special inputs
|
||||
/* Special inputs
|
||||
* processing loop
|
||||
*/
|
||||
|
||||
@ -194,7 +193,7 @@ L(SPECIAL_VALUES_LOOP):
|
||||
incl %r12d
|
||||
cmpl $4, %r12d
|
||||
|
||||
/* Check bits in range mask */
|
||||
/* Check bits in range mask */
|
||||
jl L(RANGEMASK_CHECK)
|
||||
# LOE rbx rbp r15 r12d r13d
|
||||
|
||||
@ -206,26 +205,26 @@ L(SPECIAL_VALUES_LOOP):
|
||||
cfi_restore(14)
|
||||
movups 48(%rsp), %xmm0
|
||||
|
||||
/* Go to exit */
|
||||
/* 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
|
||||
/* Scalar math fucntion call
|
||||
* to process special input
|
||||
*/
|
||||
|
||||
L(SCALAR_MATH_CALL):
|
||||
movl %r12d, %r14d
|
||||
movss 32(%rsp,%r14,4), %xmm0
|
||||
movss 32(%rsp, %r14, 4), %xmm0
|
||||
call exp10f@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 */
|
||||
/* Process special inputs in loop */
|
||||
jmp L(SPECIAL_VALUES_LOOP)
|
||||
# LOE rbx rbp r15 r12d r13d
|
||||
END(_ZGVbN4v_exp10f_sse4)
|
||||
@ -235,8 +234,7 @@ END(_ZGVbN4v_exp10f_sse4)
|
||||
|
||||
#ifdef __svml_sexp10_data_internal_typedef
|
||||
typedef unsigned int VUINT32;
|
||||
typedef struct
|
||||
{
|
||||
typedef struct {
|
||||
__declspec(align(16)) VUINT32 _sT[(1<<5)][1];
|
||||
__declspec(align(16)) VUINT32 _sLg2_10[4][1];
|
||||
__declspec(align(16)) VUINT32 _sShifter[4][1];
|
||||
@ -251,7 +249,7 @@ typedef struct
|
||||
} __svml_sexp10_data_internal;
|
||||
#endif
|
||||
__svml_sexp10_data_internal:
|
||||
/*== _sT ==*/
|
||||
/* _sT */
|
||||
.long 0x3f800000 // 2^( 0 /32 )
|
||||
.long 0x3f82cd87 // 2^( 1 /32 )
|
||||
.long 0x3f85aac3 // 2^( 2 /32 )
|
||||
@ -289,7 +287,7 @@ __svml_sexp10_data_internal:
|
||||
.align 16
|
||||
.long 0x4b400000, 0x4b400000, 0x4b400000, 0x4b400000 /* _sShifter) */
|
||||
.align 16
|
||||
.long 0x3c1a2000, 0x3c1a2000, 0x3c1a2000, 0x3c1a2000 /* _sInvLg2_10hi/2^K hi (24-K-7) bits*/
|
||||
.long 0x3c1a2000, 0x3c1a2000, 0x3c1a2000, 0x3c1a2000 /* _sInvLg2_10hi/2^K hi (24-K-7) bits */
|
||||
.align 16
|
||||
.long 0x341a84fc, 0x341a84fc, 0x341a84fc, 0x341a84fc /* _sInvLg2_10lo/2^K lo bits */
|
||||
// otherwise exp10(0) won't produce exact 1.0
|
||||
@ -300,12 +298,12 @@ __svml_sexp10_data_internal:
|
||||
.align 16
|
||||
.long 0x4029a8d2, 0x4029a8d2, 0x4029a8d2, 0x4029a8d2 /* _sPC2 */
|
||||
.align 16
|
||||
.long 0x0000001f, 0x0000001f, 0x0000001f, 0x0000001f /* _iIndexMask =(2^K-1)*/
|
||||
.long 0x0000001f, 0x0000001f, 0x0000001f, 0x0000001f /* _iIndexMask =(2^K-1) */
|
||||
//common
|
||||
.align 16
|
||||
.long 0x7fffffff, 0x7fffffff, 0x7fffffff, 0x7fffffff /* _iAbsMask */
|
||||
.align 16
|
||||
.long 0x4217b818, 0x4217b818, 0x4217b818, 0x4217b818 /* _iDomainRange=-log10(max_denormal=0x007fffff) RZ */
|
||||
.align 16
|
||||
.type __svml_sexp10_data_internal,@object
|
||||
.size __svml_sexp10_data_internal,.-__svml_sexp10_data_internal
|
||||
.type __svml_sexp10_data_internal, @object
|
||||
.size __svml_sexp10_data_internal, .-__svml_sexp10_data_internal
|
||||
|
||||
Loading…
Reference in New Issue
Block a user