[mirror_edk2.git] / MdePkg / Library / BaseMemoryLibSse2 / X64 / SetMem.S

#\r
# ConvertAsm.py: Automatically generated from SetMem.asm\r
#\r
#------------------------------------------------------------------------------\r
#\r
# Copyright (c) 2006, Intel Corporation\r
# All rights reserved. This program and the accompanying materials\r
# are licensed and made available under the terms and conditions of the BSD License\r
# which accompanies this distribution.  The full text of the license may be found at\r
# http://opensource.org/licenses/bsd-license.php\r
#\r
# THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,\r
# WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.\r
#\r
# Module Name:\r
#\r
#   SetMem.S\r
#\r
# Abstract:\r
#\r
#   SetMem function\r
#\r
# Notes:\r
#\r
#------------------------------------------------------------------------------\r
\r
\r
#------------------------------------------------------------------------------\r
#  VOID *\r
#  EFIAPI\r
#  InternalMemSetMem (\r
#    IN VOID   *Buffer,\r
#    IN UINTN  Count,\r
#    IN UINT8  Value\r
#    )\r
#------------------------------------------------------------------------------\r
.intel_syntax noprefix\r
.globl ASM_PFX(InternalMemSetMem)\r
ASM_PFX(InternalMemSetMem):\r
    push    rdi\r
    mov     rdi, rcx                    # rdi <- Buffer\r
    mov     al, r8b                     # al <- Value\r
    mov     r9, rdi                     # r9 <- Buffer as return value\r
    xor     rcx, rcx\r
    sub     rcx, rdi\r
    and     rcx, 15                     # rcx + rdi aligns on 16-byte boundary\r
    jz      L0\r
    cmp     rcx, rdx\r
    cmova   rcx, rdx\r
    sub     rdx, rcx\r
    rep     stosb\r
L0:\r
    mov     rcx, rdx\r
    and     rdx, 15\r
    shr     rcx, 4\r
    jz      L_SetBytes\r
    mov     ah, al                      # ax <- Value repeats twice\r
    movdqa  [rsp + 0x10], xmm0          # save xmm0\r
    movd    xmm0, eax                   # xmm0[0..16] <- Value repeats twice\r
    pshuflw xmm0, xmm0, 0               # xmm0[0..63] <- Value repeats 8 times\r
    movlhps xmm0, xmm0                  # xmm0 <- Value repeats 16 times\r
L1:\r
    movntdq [rdi], xmm0                 # rdi should be 16-byte aligned\r
    add     rdi, 16\r
    loop    L1\r
    mfence\r
    movdqa  xmm0, [rsp + 0x10]          # restore xmm0\r
L_SetBytes:\r
    mov     ecx, edx                    # high 32 bits of rcx are always zero\r
    rep     stosb\r
    mov     rax, r9                     # rax <- Return value\r
    pop     rdi\r
    ret\r
Commit	Line	Data
b1ff428c	1	#\r
	2	# ConvertAsm.py: Automatically generated from SetMem.asm\r
	3	#\r
	4	#------------------------------------------------------------------------------\r
	5	#\r
	6	# Copyright (c) 2006, Intel Corporation\r
	7	# All rights reserved. This program and the accompanying materials\r
	8	# are licensed and made available under the terms and conditions of the BSD License\r
	9	# which accompanies this distribution. The full text of the license may be found at\r
	10	# http://opensource.org/licenses/bsd-license.php\r
	11	#\r
	12	# THE PROGRAM IS DISTRIBUTED UNDER THE BSD LICENSE ON AN "AS IS" BASIS,\r
	13	# WITHOUT WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESS OR IMPLIED.\r
	14	#\r
	15	# Module Name:\r
	16	#\r
	17	# SetMem.S\r
	18	#\r
	19	# Abstract:\r
	20	#\r
	21	# SetMem function\r
	22	#\r
	23	# Notes:\r
	24	#\r
	25	#------------------------------------------------------------------------------\r
	26	\r
	27	\r
	28	#------------------------------------------------------------------------------\r
	29	# VOID *\r
	30	# EFIAPI\r
	31	# InternalMemSetMem (\r
	32	# IN VOID *Buffer,\r
	33	# IN UINTN Count,\r
	34	# IN UINT8 Value\r
	35	# )\r
	36	#------------------------------------------------------------------------------\r
	37	.intel_syntax noprefix\r
	38	.globl ASM_PFX(InternalMemSetMem)\r
	39	ASM_PFX(InternalMemSetMem):\r
	40	push rdi\r
	41	mov rdi, rcx # rdi <- Buffer\r
	42	mov al, r8b # al <- Value\r
	43	mov r9, rdi # r9 <- Buffer as return value\r
	44	xor rcx, rcx\r
	45	sub rcx, rdi\r
	46	and rcx, 15 # rcx + rdi aligns on 16-byte boundary\r
	47	jz L0\r
	48	cmp rcx, rdx\r
	49	cmova rcx, rdx\r
	50	sub rdx, rcx\r
	51	rep stosb\r
	52	L0:\r
	53	mov rcx, rdx\r
	54	and rdx, 15\r
	55	shr rcx, 4\r
	56	jz L_SetBytes\r
	57	mov ah, al # ax <- Value repeats twice\r
	58	movdqa [rsp + 0x10], xmm0 # save xmm0\r
	59	movd xmm0, eax # xmm0[0..16] <- Value repeats twice\r
	60	pshuflw xmm0, xmm0, 0 # xmm0[0..63] <- Value repeats 8 times\r
	61	movlhps xmm0, xmm0 # xmm0 <- Value repeats 16 times\r
	62	L1:\r
	63	movntdq [rdi], xmm0 # rdi should be 16-byte aligned\r
	64	add rdi, 16\r
65	loop L1\r
66	mfence\r
67	movdqa xmm0, [rsp + 0x10] # restore xmm0\r
68	L_SetBytes:\r
69	mov ecx, edx # high 32 bits of rcx are always zero\r
70	rep stosb\r
71	mov rax, r9 # rax <- Return value\r
72	pop rdi\r
73	ret\r