Assembly 来自AT&;的英特尔汇编ljmp语法;T语法
我正在尝试将xv6引导代码从At&t语法转换为Intel语法,但ljmp指令有问题。我正在努力学习英特尔计算机的引导过程,但我对英特尔汇编不是特别在行 原始AT&T语法是Assembly 来自AT&;的英特尔汇编ljmp语法;T语法,assembly,x86-16,gnu-assembler,att,intel-syntax,Assembly,X86 16,Gnu Assembler,Att,Intel Syntax,我正在尝试将xv6引导代码从At&t语法转换为Intel语法,但ljmp指令有问题。我正在努力学习英特尔计算机的引导过程,但我对英特尔汇编不是特别在行 原始AT&T语法是ljmp$0x8,$start32 最简单的例子: .code16 jmp 0x8:start32 # won't assemble .code32 start32: nop 将as-32-msyntax=intel-mnaked reg foo.s与GNU Binutils 2.35.1一起使
ljmp$0x8,$start32
最简单的例子:
.code16
jmp 0x8:start32 # won't assemble
.code32
start32:
nop
将as-32-msyntax=intel-mnaked reg foo.s
与GNU Binutils 2.35.1一起使用会产生Error:junk:far jmp行的表达式后的start32
我正在使用GNU as和gcc工具。该程序集还可能存在其他问题,例如gdtdesc和gdt 移植到英特尔语法的完整代码如下:
# Start the first CPU: switch to 32-bit protectied mode, jump into C.
# The BIOS loads this code from the first sector of the hard disk into
# memory at physical address 0x7c00 and starts executing in real mode
# with cs = 0 and ip = 7c00.
.code16
.global start
start:
# Disable interrupts.
cli
# Zero data segment registers DS, ES, and SS.
xor ax, ax
mov ds, ax
mov es, ax
mov ss, ax
seta20.1:
# Wait for not busy.
in al, 0x64
test al, 0x2
jnz seta20.1
# 0xd1 -> port 0x64
mov al, 0xd1
out 0x64, al
seta20.2:
# Wait for not busy.
in al, 0x64
test al, 0x2
jnz seta20.2
# 0xdf -> port 0x60
mov al, 0xdf
out 0x60, al
# Switch from real to protected mode. Use a bootstrap GDT that makes
# virtual addresses map directly to physical addressses so that the
# effective memory map doesn't change during the transition.
lgdt gdtdesc
# Protection Enable in cr0 register.
mov eax, cr0
or eax, 0x1
mov cr0, eax
# Complete the transtion to 32-bit protected mode by using a long jmp
# to reload cs and eip. The segment descriptors are set up with no
# translation, so that the mapping is still the identity mapping.
# This instruction giving me problems.
ljmp start32, 0x8
.code32
start32:
# Set up the protected-mode data segment registers
mov ax, 0x10
mov ds, ax
mov es, ax
mov ss, ax
# Zero the segments not ready for use.
xor ax, ax
mov fs, ax
mov gs, ax
# Set up the stack pointer and call into C.
mov esp, start
call bootmain
# If bootmain returns spin.. ??
spin:
hlt
jmp spin
# Bootstrap GDT set up null segment, code segment, and data segment respectively.
# Force 4 byte alignment.
.p2align 2
gdt:
.word 0x0000, 0x0000
.byte 0, 0, 0, 0
.word 0xffff, 0x0000
.byte 0, 0x9a, 0xcf, 0
.word 0xffff, 0x0000
.byte 0, 0x92, 0xcf, 0
# sizeof(gdt) - 1 and address of gdt respectively.
gdtdesc:
.word (gdtdesc - gdt - 1)
.long gdt
ljmp 0x08, start32
在您提供的完整翻译代码中,此行不正确:
ljmp start32, 0x8
GNU汇编程序英特尔语法中FAR JMP的正确语法为:
# Start the first CPU: switch to 32-bit protectied mode, jump into C.
# The BIOS loads this code from the first sector of the hard disk into
# memory at physical address 0x7c00 and starts executing in real mode
# with cs = 0 and ip = 7c00.
.code16
.global start
start:
# Disable interrupts.
cli
# Zero data segment registers DS, ES, and SS.
xor ax, ax
mov ds, ax
mov es, ax
mov ss, ax
seta20.1:
# Wait for not busy.
in al, 0x64
test al, 0x2
jnz seta20.1
# 0xd1 -> port 0x64
mov al, 0xd1
out 0x64, al
seta20.2:
# Wait for not busy.
in al, 0x64
test al, 0x2
jnz seta20.2
# 0xdf -> port 0x60
mov al, 0xdf
out 0x60, al
# Switch from real to protected mode. Use a bootstrap GDT that makes
# virtual addresses map directly to physical addressses so that the
# effective memory map doesn't change during the transition.
lgdt gdtdesc
# Protection Enable in cr0 register.
mov eax, cr0
or eax, 0x1
mov cr0, eax
# Complete the transtion to 32-bit protected mode by using a long jmp
# to reload cs and eip. The segment descriptors are set up with no
# translation, so that the mapping is still the identity mapping.
# This instruction giving me problems.
ljmp start32, 0x8
.code32
start32:
# Set up the protected-mode data segment registers
mov ax, 0x10
mov ds, ax
mov es, ax
mov ss, ax
# Zero the segments not ready for use.
xor ax, ax
mov fs, ax
mov gs, ax
# Set up the stack pointer and call into C.
mov esp, start
call bootmain
# If bootmain returns spin.. ??
spin:
hlt
jmp spin
# Bootstrap GDT set up null segment, code segment, and data segment respectively.
# Force 4 byte alignment.
.p2align 2
gdt:
.word 0x0000, 0x0000
.byte 0, 0, 0, 0
.word 0xffff, 0x0000
.byte 0, 0x9a, 0xcf, 0
.word 0xffff, 0x0000
.byte 0, 0x92, 0xcf, 0
# sizeof(gdt) - 1 and address of gdt respectively.
gdtdesc:
.word (gdtdesc - gdt - 1)
.long gdt
ljmp 0x08, start32
选择器值为第一个,偏移量为第二个。从AT&T语法的翻译中,您似乎颠倒了这两个值,而顺序本应保持不变。如果将这些值反转,则会出现错误error:cannothandlenon-absolutesegment in'ljmp'
。在GNU汇编程序的英特尔语法中,您还可以用jmp
替换ljmp
,这样jmp 0x08,start32
也可以工作
英特尔语法有不同的风格jmp 0x8:start32
是NASM的英特尔语法,它不同于GNU汇编程序的英特尔语法,后者的:
和,
不同。如果使用:
来分隔这两个值,则会在GNU汇编程序中的表达式后得到错误error:junk':start32'
笔记
- 如果
bootmain
中的代码不起作用,则可能是与您在本问题中介绍的引导加载程序代码无关的问题。如果您还使用Intel语法而不是AT&T语法构建所有C代码,那么请确保所有内联程序集都已正确转换,因为源代码和操作数也将被反转。xv6可能在许多文件中都有内联程序集,包括xv6 public/x86.h
,xv6 public/spinlock.c
,xv6 public/usertests.c
和xv6 public/stressfs.c
您可以使用jmp 0x08,start32
由于某些原因,jmp 0x8:start32
仅在之后才起作用。英特尔语法noprefix
,即使命令行参数应该是等效的。这是Binutils使用的语法objdump-d-Mintel-mi8086
,例如ea 16 00 08 00 jmp 0x8:0x16
,因此它可能是一个气体错误,有时不被接受
根据您对Jester的回复,我编辑了您的问题,创建了一个小的可复制示例,使用as
2.35.1(我在Arch GNU/Linux上有这个示例)。我包括了命令行选项:我假设您一定使用了这些选项,因为您的文件中没有.intel\u syntax noprefix
指令
这似乎是问题所在:-msyntax=intel-mnaked reg
使其他英特尔语法功能正常工作,如xor ax,ax
,但不能使jmp 0x8:start32
正常工作(或以其他方式编写)。只有.intel_syntax noprefix
1指令才能使far jmp的语法工作
# .intel_syntax noprefix # rely on command line options to set this
.code16
xor ax, ax # verify that command-line setting of intel_syntax worked, otherwise this line errors.
ljmp 0x8, start32 # Working before or after a syntax directive, but is basically AT&T syntax
# jmp 0x8:start32 # fails here, works after a directive
jmp 0x8, start32 # Michael Petch's suggested syntax that's still somewhat AT&Tish. works with just cmdline opts.
.att_syntax
ljmp $0x8, $start32 # working everywhere, even with clang
.intel_syntax noprefix
jmp 0x8:start32 # objdump disassembly syntax, but only works after a .intel_syntax noprefix directive
.code32
start32:
nop
我验证了-msyntax=intel-mnaked reg
适用于其他需要其效果的指令:movzx ax,al
有效。但是如果没有-mnaked reg
,我们将得到“太多内存引用”,因为“ax”和“al”将被视为符号名。没有或“操作数大小不匹配”没有-msyntax=intel
A GAS listfromas-32-msyntax=intel-mmnemonic=intel-mnaked reg-o foo.o foo.s-al--list lhs width=2--list rhs width=140
(我非常确定-mmnemonic=intel
是不相关的,syntax=intel暗示了这一点。)
注意,您可以看到哪些指令工作是因为它们有机器代码,而哪些指令没有工作(第一个jmp 0x8:start32
),因为左边的列为空。第一列通常是地址,但是是????因为程序集失败。(因为我取消了jmp 0x8:start32的注释,以显示它第一次失败,第二次工作。)
(GAS没有在“words”中列出左列的字段宽度,这显然意味着32位块。这就是为什么段选择器的00
最高有效字节被空格分隔的原因。)
在jmp 0x8:label
之前放置标签没有帮助;这不是向前或向后参考的问题。即使jmp 0x8:23
也无法组装
反汇编程序从工作版本中“推荐”的语法:
objdump-drwC-Mintel-mi8086 foo.o
:
foo.o: file format elf32-i386
Disassembly of section .text:
00000000 <start32-0x17>:
0: 0f b6 c0 movzx ax,al
3: ea 17 00 08 00 jmp 0x8:0x17 4: R_386_16 .text
8: ea 17 00 08 00 jmp 0x8:0x17 9: R_386_16 .text
d: ea 17 00 08 00 jmp 0x8:0x17 e: R_386_16 .text
12: ea 17 00 08 00 jmp 0x8:0x17 13: R_386_16 .text
00000017 <start32>:
17: 90 nop
00000000 <.text>:
0: 0f b6 c0 movzx ax, al
3: ea 17 00 08 00 ljmp 8, 23
8: ea 17 00 08 00 ljmp 8, 23
d: ea 17 00 08 00 ljmp 8, 23
12: ea 17 00 08 00 ljmp 8, 23
00000017 <start32>:
17: 90 nop
顺便说一句,我没有让Clang11.0用符号名来组装任何英特尔语法版本LJMP8,12
以铿锵声进行汇编,但甚至不LJMP8,start32
。只有切换到AT&T语法并返回,我才能让clang的内置汇编程序(clang-m32-masm=intel-c
)发出16位模式的远jmp
.att_syntax
ljmp $0x8, $start32 # working everywhere, even with clang
.intel_syntax noprefix
请记住,这种直接形式的far JMP在64位模式下不可用;也许这就是为什么LLVM的内置汇编器在这方面花费更少的精力
脚注1:实际上,。英特尔语法前缀
也可以,但永远不要使用它。没有人想看到弗兰肯怪物是mov%eax,[%eax]
,尤其是add%edx,%eax
,它使用的是dst,src
顺序,但有AT&T修饰的寄存器名。你想要jmp 0x8:start32
。PS:atat是《星球大战行者》,语法是at&t:)我的坏哈哈。使用jmp 0x8:start32会在表达式“Using GNU as.Assembly”之后生成“Error:junk”:start32,在这里使用as
version 2.28可以很好地进行汇编。你有什么