OSDev.org

Thanks for reply.

You don't sound rude. I know I don't know asm very well yet. But I will learn it. I don't belive that one guy in the forum who told he had learned asm basics in 4 hours.

But could you still tell me what is SELECTORkernelCode for I could use the function? (I think I understand the rest of it in some way?!?!)

I don't want to interfere here, but I *guess* it is a constant that contains the selector of the code segment which contains the kernel code. From looking at the code, this seems to make sense.

cheers Joe

Hi again!

When I have this in data section:
;;;;;;;;;;;;;;;;;;;;;;;;
idt_start:
dw 0x0000 ; <--<
dw 0x08
dw 0x8E00
dw 0x20 ; <--<
...and so on for all of the slots...
idt_end:
;;;;;;;;;;;;;;;;;;;;;;;;
..., how can I point to the "arrow places" from the asm code? I tried :
mov word [idt_start+1], eax
and
mov word [idt_start+4], eax
but I guess it doesn't work!?

(idt_start and -_end are global symbols)

AL/AH = 1byte, AX=2bytes, EAX=4bytes [and RAX=8bytes]

A word (DW) is 2 bytes so idt_start + 2*(entry-1).

Thanks.

I wrote this by myself :)

mov ax, isr0-$$ & 0xFFFF
mov [start_of_idt],ax
mov ax, (isr0-$$ >> 16) & 0xFFFF
mov [start_of_idt+6],ax

I meant it to fill the first slot of the empty IDT with isr #0 (, which exists of course), but it crashes my kernel! Why?
(I call it before "sti" and "call main")

Please tell me the correct version of this.

Please answer the message ^ above ^ ; otherways I'm stuck in OS-dev.!

One question:
Excuse me if I missed something, but: why don't you code the idt in c?

Because I think I can get it more clear in asm; like if I'll want to port my os to a new arch some day I can just rewrite the asm.

I want to do everything like IDT, GDT and so-on in asm-file. C files can then contain everything related with the real work; asm for initialization and the base.

So, could you help me with my problem?!

(Sorry, the code above won't crash my kernel: it just does nothing. What's wrong with it?)

So, will you help me?

Why "-$$"?
Why not hardcoding it (you can still change it at runtime later)?

If I leave "-$$" off it will give me syntax error, something like "operator may only be applied to scalar values"

Is isr0-$$ a pointer to isr0 label?

Please tell me why this code doesn't work:
(That code was supposed to fill slot#0 of IDT with isr0's address)

Before that code I do LIDT. After it I call main. And after calling main I have:
And in C I have:


But interrupt_service_routine0 never gets called! Why?
(if I do STI before calling main: kernel crashes)
(if I do asm("sti"); nothing happens)

If you have only implemented ISR0, ISR0=Divide by Zero Exception, any other interrupt (ie. the periodic PIT hardware interrupt) will fail which will double fault which isn't handled either which will triple fault and reboot.

If you have only implemented ISR0...

No, I have not.

As I told you, there was something wrong with my code, and I'd like that if you made an exception and told me what's wrong with it.

I tried to divide n/0 in main but nothing happened. Did you mean I have to implement some other interrupts to get any of them working?

BTW do you mean IRQ0 instead interrupt 0? Because if you are expecting to catch the timer interrupt that is called periodically, you need to set up the PIC. First you need to re-map the IRQ (hardware) interupts so that they don't conflict with various protected mode stuffs. Then you also need to unmask the IRQ's you want to use. There is a tutorial about this on bonafide I believe.

Also, I generally find c more portable, as with a change in instruction sets you only have to get a compiler for that architechture, whereas with asm you might end up replacing all the instructions manually!

Perhaps you're right with that portable-code thing! I mean the interrupt 0 which is described at the first entry of IDT. I will do it in C later, but could you now PLEASE give me what I've asked for in 6 mails already!?!!!!!!