the jump form rmode to pmode

[faway]

Code: Select all

bits 16
org 0x7e00
jmp start
 	%include "stdio.inc"
	%include "gdt.inc"
	%include "A20.inc"

start:
	cli
	xor ax,ax
	mov ds,ax
	mov es,ax
	mov ax,0x9000
	mov ss,ax
	mov sp,0xffff

	call loadGDT
	
	call enableA20
	
	;cli
	
	mov eax,cr0
	or eax,1
	mov cr0,eax
	
	[color=#BF0000]jmp code_desc:protected_code[/color]
[bits 32]
protected_code:
	......

	times 1024-($-$$) db 0

This is the switch code about it.Before this code,i has done a lot of work in in the real mode.When bochs got the jmp,some error happend:
Error: (0) print_guard_results: guard_found ? (stop reason 0)
However,if i delete the job which in the real mode ,it ok.

[faway]

Code: Select all

[bits 16]
loadGDT:
	cli
	pusha
	lgdt [gdt_desc]
	popa
	ret
gdt:
	dd 0
	dd 0
code_desc equ $-gdt
gdt_code:
	dw 0xffff
	dw 0
	db 0
	db 10011010b
	db 11001111b
	db 0
gdt_data:
	dw 0xffff
	dw 0
	db 0
	db 10010010b
	db 11001111b
	db 0
gdt_end:
gdt_desc:
	dw gdt_end - gdt  - 1
	dd gdt

this is the code about GDT.
Please help me.

[Love4Boobies]

Your origin is 7E00h.

[faway]

I am sorry "Love4Boobies ",what's wrong about 0x7e00,this is the adress where I load the pmode code,and rmode is in the other place.
it's not the boot file.

[Love4Boobies]

Ah, ok. So what is it that you do in real mode? Maybe you do something there that screwes up the rest (like writing over the EBDA or something).
EDIT: and are you sure this: jmp code_desc:protected_code is right?

[faway]

Ah, ok. So what is it that you do in real mode? Maybe you do something there that screwes up the rest (like writing over the EBDA or something).
EDIT: and are you sure this: jmp code_desc:protected_code is right?

The "code_desc" has define in the second code about GDT.
Probably you are right,something may be writted,but I don't know how to check and amend.
I use bochs to debug,and I got the some info about registers just as follow.
before jmp:

Code: Select all

<bochs:9> info r
eax            0x60000011       1610612753
ecx            0x130000         1245184
edx            0x0              0
ebx            0x2              2
esp            0xffff           0xffff
ebp            0x0              0x0
esi            0x0              0
edi            0x10a            266
eip            0x1d3            0x1d3
eflags         0x46             70
cs             0x7e00           32256
ss             0x9000           36864
ds             0x40             64
es             0x0              0
fs             0x0              0
gs             0x0              0

after jmp

Code: Select all

<bochs:11> info r
eax            0x0              0
ecx            0x0              0
edx            0x300            768
ebx            0x0              0
esp            0x0              0x0
ebp            0x0              0x0
esi            0x0              0
edi            0x0              0
eip            0xfff0           0xfff0
eflags         0x2              2
cs             0xf000           61440
ss             0x0              0
ds             0x0              0
es             0x0              0
fs             0x0              0
gs             0x0              0

[Love4Boobies]

Ah, ok. So what is it that you do in real mode? Maybe you do something there that screwes up the rest (like writing over the EBDA or something).
EDIT: and are you sure this: jmp code_desc:protected_code is right?

The "code_desc" has define in the second code about GDT.

Yes. code_desc is a constant whose value is 8. Are you sure that's what you want? Are you sure you don't mean jmp protected_code instead? The code in you file is in the same segment so there's no need for a far jump.

[Combuster]

um, the far jump is needed to change from 16 bit code to 32 bit code

[Love4Boobies]

um, the far jump is needed to change from 16 bit code to 32 bit code

Ooops. Right, he was jumping to pmode. Still, be careful with the segment.

[faway]

um, the far jump is needed to change from 16 bit code to 32 bit code

Ooops. Right, he was jumping to pmode. Still, be careful with the segment.

If I code just as "jmp protected_code", it can jmp to the 32bit.However,when I init registers,the same error has happened. "mov es,ax" can't be executd.

Code: Select all

	jmp protected_code
[bits 32]
protected_code:
	  mov ax , data_desc  
    mov ds , ax
    [color=#FF0000]mov es , ax[/color]
    mov ss , ax 
    mov fs , ax 
    mov gs , ax
    mov esp , 0xffff 

[faway]

The problem is I have no idea how to manage the space if it complicated.

Code: Select all

Error: (0) print_guard_results: guard_found ? (stop reason 0)

It is driving me crazy!

[Troy Martin]

One word: Google.

Another word: Wiki.

A final word: http://www.catb.org/~esr/faqs/smart-questions.html

[Combuster]

One word: Google.

Query text wanted. Plugging the error message yields no usable results.

EDIT: Have you fixed the ORG problem? could you post an updated register dump on the state before the far jump? From what I gathered the error message looks like a reset (in your case, triple fault).

[frank]

The problem is that CS isn't zero. The code goes from running at 0x7E00:something to 0:same something. The new GDT has a base of 0 and the old CS has a base of 0x7E00. I don't know how the NASM org instruction works exactly so I can't help you any further.

EDIT: Is the code running at 0x7E00 or 0x7E000? CS suggests the second and the ORG instruction the other.

[Troy Martin]

One word: Google.

Query text wanted. Plugging the error message yields no usable results.

I'm putting my invisible tokens on the line by saying that I bet it's a problem with the GDT.

cs 0xf000 61440

Something here looks incorrectly set up. Probably either the GDT or the jump. Probably also causing a triple fault.

OP: try, instead of using your code_desc and data_desc things, use 0x08 and 0x10 respectively.