[ceph.git] / ceph / src / arrow / go / arrow / math / _lib / int64_sse4.s

	.text
	.intel_syntax noprefix
	.file	"_lib/int64.c"
	.globl	sum_int64_sse4
	.p2align	4, 0x90
	.type	sum_int64_sse4,@function
sum_int64_sse4:                         # @sum_int64_sse4
# BB#0:
	push	rbp
	mov	rbp, rsp
	and	rsp, -8
	test	rsi, rsi
	je	.LBB0_1
# BB#2:
	cmp	rsi, 3
	jbe	.LBB0_3
# BB#6:
	mov	r9, rsi
	and	r9, -4
	je	.LBB0_3
# BB#7:
	lea	r8, [r9 - 4]
	mov	eax, r8d
	shr	eax, 2
	inc	eax
	and	rax, 3
	je	.LBB0_8
# BB#9:
	neg	rax
	pxor	xmm0, xmm0
	xor	ecx, ecx
	pxor	xmm1, xmm1
	.p2align	4, 0x90
.LBB0_10:                               # =>This Inner Loop Header: Depth=1
	movdqu	xmm2, xmmword ptr [rdi + 8*rcx]
	movdqu	xmm3, xmmword ptr [rdi + 8*rcx + 16]
	paddq	xmm0, xmm2
	paddq	xmm1, xmm3
	add	rcx, 4
	inc	rax
	jne	.LBB0_10
	jmp	.LBB0_11
.LBB0_3:
	xor	r9d, r9d
	xor	eax, eax
.LBB0_4:
	lea	rcx, [rdi + 8*r9]
	sub	rsi, r9
	.p2align	4, 0x90
.LBB0_5:                                # =>This Inner Loop Header: Depth=1
	add	rax, qword ptr [rcx]
	add	rcx, 8
	dec	rsi
	jne	.LBB0_5
	jmp	.LBB0_15
.LBB0_1:
	xor	eax, eax
.LBB0_15:
	mov	qword ptr [rdx], rax
	mov	rsp, rbp
	pop	rbp
	ret
.LBB0_8:
	xor	ecx, ecx
	pxor	xmm0, xmm0
	pxor	xmm1, xmm1
.LBB0_11:
	cmp	r8, 12
	jb	.LBB0_14
# BB#12:
	mov	rax, r9
	sub	rax, rcx
	lea	rcx, [rdi + 8*rcx + 112]
	.p2align	4, 0x90
.LBB0_13:                               # =>This Inner Loop Header: Depth=1
	movdqu	xmm2, xmmword ptr [rcx - 112]
	movdqu	xmm3, xmmword ptr [rcx - 96]
	movdqu	xmm4, xmmword ptr [rcx - 80]
	movdqu	xmm5, xmmword ptr [rcx - 64]
	paddq	xmm2, xmm0
	paddq	xmm3, xmm1
	movdqu	xmm6, xmmword ptr [rcx - 48]
	movdqu	xmm7, xmmword ptr [rcx - 32]
	paddq	xmm6, xmm4
	paddq	xmm6, xmm2
	paddq	xmm7, xmm5
	paddq	xmm7, xmm3
	movdqu	xmm0, xmmword ptr [rcx - 16]
	movdqu	xmm1, xmmword ptr [rcx]
	paddq	xmm0, xmm6
	paddq	xmm1, xmm7
	sub	rcx, -128
	add	rax, -16
	jne	.LBB0_13
.LBB0_14:
	paddq	xmm0, xmm1
	pshufd	xmm1, xmm0, 78          # xmm1 = xmm0[2,3,0,1]
	paddq	xmm1, xmm0
	movq	rax, xmm1
	cmp	r9, rsi
	jne	.LBB0_4
	jmp	.LBB0_15
.Lfunc_end0:
	.size	sum_int64_sse4, .Lfunc_end0-sum_int64_sse4


	.ident	"Apple LLVM version 9.0.0 (clang-900.0.39.2)"
	.section	".note.GNU-stack","",@progbits
Commit	Line	Data
1d09f67e TL	1	.text
	2	.intel_syntax noprefix
	3	.file "_lib/int64.c"
	4	.globl sum_int64_sse4
	5	.p2align 4, 0x90
	6	.type sum_int64_sse4,@function
	7	sum_int64_sse4: # @sum_int64_sse4
	8	# BB#0:
	9	push rbp
	10	mov rbp, rsp
	11	and rsp, -8
	12	test rsi, rsi
	13	je .LBB0_1
	14	# BB#2:
	15	cmp rsi, 3
	16	jbe .LBB0_3
	17	# BB#6:
	18	mov r9, rsi
	19	and r9, -4
	20	je .LBB0_3
	21	# BB#7:
	22	lea r8, [r9 - 4]
	23	mov eax, r8d
	24	shr eax, 2
	25	inc eax
	26	and rax, 3
	27	je .LBB0_8
	28	# BB#9:
	29	neg rax
	30	pxor xmm0, xmm0
	31	xor ecx, ecx
	32	pxor xmm1, xmm1
	33	.p2align 4, 0x90
	34	.LBB0_10: # =>This Inner Loop Header: Depth=1
	35	movdqu xmm2, xmmword ptr [rdi + 8*rcx]
	36	movdqu xmm3, xmmword ptr [rdi + 8*rcx + 16]
	37	paddq xmm0, xmm2
	38	paddq xmm1, xmm3
	39	add rcx, 4
	40	inc rax
	41	jne .LBB0_10
	42	jmp .LBB0_11
	43	.LBB0_3:
	44	xor r9d, r9d
	45	xor eax, eax
	46	.LBB0_4:
	47	lea rcx, [rdi + 8*r9]
	48	sub rsi, r9
	49	.p2align 4, 0x90
	50	.LBB0_5: # =>This Inner Loop Header: Depth=1
	51	add rax, qword ptr [rcx]
	52	add rcx, 8
	53	dec rsi
	54	jne .LBB0_5
	55	jmp .LBB0_15
	56	.LBB0_1:
	57	xor eax, eax
	58	.LBB0_15:
	59	mov qword ptr [rdx], rax
	60	mov rsp, rbp
	61	pop rbp
	62	ret
	63	.LBB0_8:
	64	xor ecx, ecx
65	pxor xmm0, xmm0
66	pxor xmm1, xmm1
67	.LBB0_11:
68	cmp r8, 12
69	jb .LBB0_14
70	# BB#12:
71	mov rax, r9
72	sub rax, rcx
73	lea rcx, [rdi + 8*rcx + 112]
74	.p2align 4, 0x90
75	.LBB0_13: # =>This Inner Loop Header: Depth=1
76	movdqu xmm2, xmmword ptr [rcx - 112]
77	movdqu xmm3, xmmword ptr [rcx - 96]
78	movdqu xmm4, xmmword ptr [rcx - 80]
79	movdqu xmm5, xmmword ptr [rcx - 64]
80	paddq xmm2, xmm0
81	paddq xmm3, xmm1
82	movdqu xmm6, xmmword ptr [rcx - 48]
83	movdqu xmm7, xmmword ptr [rcx - 32]
84	paddq xmm6, xmm4
85	paddq xmm6, xmm2
86	paddq xmm7, xmm5
87	paddq xmm7, xmm3
88	movdqu xmm0, xmmword ptr [rcx - 16]
89	movdqu xmm1, xmmword ptr [rcx]
90	paddq xmm0, xmm6
91	paddq xmm1, xmm7
92	sub rcx, -128
93	add rax, -16
94	jne .LBB0_13
95	.LBB0_14:
96	paddq xmm0, xmm1
97	pshufd xmm1, xmm0, 78 # xmm1 = xmm0[2,3,0,1]
98	paddq xmm1, xmm0
99	movq rax, xmm1
100	cmp r9, rsi
101	jne .LBB0_4
102	jmp .LBB0_15
103	.Lfunc_end0:
104	.size sum_int64_sse4, .Lfunc_end0-sum_int64_sse4
105
106
107	.ident "Apple LLVM version 9.0.0 (clang-900.0.39.2)"
108	.section ".note.GNU-stack","",@progbits