So I have ran into a problem which I've been stuck on now for what seems like eternity, about 3 days without sleep trying to solve the issue on my own now, and it's kinda depressing to me as I feel as if it should honestly be a relatively simple task to load a simple table and perform a far jump into 32-bit mode and boom! Have it working with the snap of my figures. The problem is causing my computer to reboot every time I load the GDT and perform a far jump however, as I have seen with numerous other posts and people dealing with the same issue. I am loading from my bootloader using 'INT 0x13, AX = 0x20A' with 'ES:BX = 0x1000:0'. Right before the 'CLI LGDT [GDT_END]' instruction, if I perform a 'JMP $' instruction, it will successfully print the string about having made it to the kernel. However, the second I load the GDT, whether loading via an %include or pasting the include's contents directly within the kernel itself, the error persists in an endless reboot.
Any suggestions in what I may be doing wrong would surely be lovely! Days spent and I can't seem to find and answer in resolving my issue anywhere, despite tons of posts seeming to resemble the same issue. Here is the kernel code I have typed out over and over again probably several thousands of times already... I tried making it look as legible as I possibly can, and tried all solutions I could find that seemed to help others. This seems to be my last resort.. If the bootloader is needed I can post it as well, however, like I said, it does in fact make it to the kernel loaded in at segment: 0x1000 offset: 0. Which I usually never use ORG, and honestly have no idea what is actually does or is doing, nor can I find clear explanations on it either, but with or without it, either way, it doesn't work. Assembled with NASM in windows cmd line using 'nasm -okernel.bin kernel.asm'.
BITS 16
ORG 0x10000
JMP FLOW_KERNEL_START
NOP
;----------------------------
; GLOBAL DESCRIPTOR TABLE
;----------------------------
GDT_START:
dq 0
GDT_CODE_SEGMENT:
dw 0xFFFF
dw 0
db 0
db 10011010b
db 11001111b
db 0
GDT_DATA_SEGMENT:
dw 0xFFFF
dw 0
db 0
db 10010010b
db 11001111b
db 0
GDT_END:
dw GDT_END - GDT_START - 1
dd GDT_START
;----------------------------
; INCLUDES
;----------------------------
;%INCLUDE 'BIO_Land\gdt.asm'
;----------------------------
FLOW_KERNEL_START:
; SET SEGMENTS AND BOOT DEVICE VARIABLES
XOR AX, AX
CLI
MOV SS, AX
MOV SP, 0x9000
STI
MOV AX, 0x1000
MOV DS, AX
MOV ES, AX
MOV FS, AX
MOV GS, AX
MOV byte [VAR_BOOTDEV], DL
CALL FUNC_GOOD
MOV AX, 0
INT 0x16
; LOAD GDT TABLE AND MASK 32-BIT PROTECTED MODE FLAG
CLI
LGDT [GDT_END]
MOV EAX, 0x10
MOV DS, EAX
MOV ES, EAX
MOV FS, EAX
MOV GS, EAX
MOV SS, EAX
MOV EAX, CR0
OR EAX, 1
MOV CR0, EAX
JMP dword 0x8:FLOW_PROTECTED_MODE
NOP
BITS 32
FLOW_PROTECTED_MODE:
STI
MOV byte [0xB8000], 'O'
MOV byte [0xB8001], 0x0F
MOV byte [0xB8002], 'O'
MOV byte [0xB8003], 0x0F
MOV byte [0xB8004], 'O'
MOV byte [0xB8005], 0x0F
MOV byte [0xB8006], 'O'
MOV byte [0xB8007], 0x0F
MOV byte [0xB8008], 'O'
MOV byte [0xB8009], 0x0F
JMP $
;----------------------------
; FUNCTIONS
;----------------------------
FUNC_GOOD:
MOV SI, VAR_STRING
CALL FUNC_PRINT
RET
FUNC_PRINT:
MOV AH, 0x0E
FUNC_PRINT_LOOP:
LODSB
CMP AL, 0
JE FUNC_PRINT_DONE
INT 0x10
JMP FUNC_PRINT_LOOP
FUNC_PRINT_DONE:
RET
;----------------------------
; VARIABLES
;----------------------------
VAR_BOOTDEV db 0
;----------------------------
; STRINGS
;----------------------------
VAR_STRING db 'Hello world! You made it to the kernel!', 0
EDIT: I have changed the the JMP $ to an INT 0x16 to make things more simple. It boots to it just fine as mentioned before but on keypress is a reboot. I have changed the ORG to 0x1000 as well as 0x10000. I'm unsure on how I would load at address 0x7C00 without attaching it to being inside the bootloader itself, but that being said I have also tried putting a buffer label at the end of the bootloader and loading the kernel at it, assuming the address after the bootloader being 0x7E00 and using it as the ORG. I have also tried omitting the ORG entirely (as it is already being set via the DS is it not?) and then changing the 0x8:FLOW_PROTECTED_MODE to 0x8:FLOW_PROTECTED_MODE+0x10000. I have also tried using JMP 0x8:$$+$ among numerous other things. BOSHC is saying exactly what I was expecting to find that the JMPF 0x8:0x00001065 cannot be resolved. I am assuming that the kernel IS in fact loaded somewhere within memory as it does say Hello world!, but I just cannot for the life of me figure out how to find the location correctly or what the problem is.
