Boch error...

[sion]

Well since my last post(not too long ago) I managed to load a simple kernel and i even tried to load a second bootloader and kernel...and to my surprise it worked so i've managed to move along steadily. But now my problem is in bochs. im programming in win95 using bochs 2.0.2 and when i try loading up the kernel it gives me the error prefetch: RIP > CS.limit ... I did a search on the board but only found EIP > CS.limit errors not RIP so i don't know....any ideas out there???

[Unspoken_Magi]

RIP is EIP, its just misspelled in Bochs. (Minor bug)

[sion]

Hmm...well that explains it...thanks...ill do a search and see if i can figure it out.

[FlashBurn]

Correct me if I?me wrong, but isn?t RIP the 64bit variant of EIP? I think AMD call teir register so.

[Tim]

Right, it's not a misspelling but part of Bochs' AA-64 emulation.

[sion]

i know this error is caused by my jump to the kernel, but is it because im loading it at 0x1000 or is it the actual jump code? ??? i just loaded it to memory and did
mov ax, 0x1000
mov es, ax
mov ds, ax
push ax
xor ax,ax
push ax
retf

maybe its the retf that bochs is not agreeing with? when i do a jmp 0x0000:0x1000 it doesn't even work at all on bochs or on a reboot....so i really don't know...i still have a lot to learn...

[FlashBurn]

Maybe you could show your loading code! Aand before you could test this jump:

Code: Select all

jmp 0x1000:0000

[Pype.Clicker]

just a question ... is the code that issue that jump in protected or real mode ?

[sion]

Ok this is my loader...

Code: Select all

BITS    16
ORG     0

        JMP short Start
        NOP

        osname             db      'Test'
        bytespersector  dw      512
        sectorspertrack dw      18
        numberofheads dw      2
        totalsectors       dw      2880
        bootdrive           db      0

Start:
        CLI        
        MOV AX, 0x7c0
        MOV DS, AX
        MOV ES, AX

        MOV AX, 0x9000
        MOV SS, AX
        MOV SP, 0xffff
        STI

        CALL Clear_Screen
        MOV SI, welcome  
        CALL Print_It

;===LOAD KERNEL===
Read_Sector:
        PUSH DS
        MOV AX, 0
        MOV DL, 0
        INT 13h
        POP DS
        JC  Kernel_Error

        PUSH ES
        MOV AX, 0x1000      ;Why exactly does this have to be here? it gets changed
        MOV ES, AX      ;when AH and AL..so is it to set the buffer for int 0x13?
        MOV BX, 0

        MOV AH, 2
        MOV AL, 5
        MOV CH, 0
        MOV CL, 2
        MOV DH, 0
        INT 13h
        POP ES
        JC Kernel_Error

        MOV AX, 0x1000      ;So now the kernel is loaded from the second sector
        MOV ES, AX      ;So why reset ES, and why set DS at all? 
        MOV DS, AX
        
        JMP 0x1000:0000
Kernel_Error:
        MOV SI, error
        CALL Print_It
        CALL Read_Sector
;===FUNTIONS===
Print_It:                
        LODSB           
        OR AL,AL
        JZ Print_Done
        MOV AH,0x0E
        MOV BX,0x07
        INT 0x10
        JMP Print_It
Print_Done:
        ret
Clear_Screen:
        MOV AL,3        
        MOV AH,0
        INT 10H
        RET
;======DATA===
        welcome db 'Bootloader...',13,10,0
        error db 'Could not read Kernel...',13,10,0

        TIMES   510-($-$$) DB 0
        dw      0xAA55

So yea...it works on reboots...but not in bochs? ??? ???

[sion]

Just for those who might be curious heres the bochsout.txt file with the error...

[sion]

...

[attachment deleted by admin]

[Unspoken_Magi]

Hmm... looks like it jumped to bogus memory somewhere....

[sion]

Well after looking at the bochs output i realized that the CS base was being set at 0009:0000 instead of 0000:9000 <---where the kernel was being loaded too....so i changed the code a bit and the 'panic:RIP > CS.Limit' error has gone...but bochs still doesn't load the kernel.. and it does load fine on a reboot...ah well i guess ill keep trying...