Page 1 of 1
kinit() not assembling under nasm
Posted: Sun Jun 15, 2003 7:00 am
by unknown user
hi. i'm having some problems getting my pmode/initial-task function to assemble. i'm using nasm, but i don't know enough about nasm to write good nasm code i guess. i'm getting wierd errors like "negative times -114 not allowed", and i didn't even make a times operator negative. can someone tell me what i'm doing wrong?
i tried to post it with the file uploaded, but it told me that a file with the same name was already uploaded, so i'm pasting into here.
Code: Select all
init:
;segment descriptor
struc segment_descriptor
.seg_length0_15 resw 0 ;low word of segment length
.base_addr0_15 resw 0 ;low word of base address
.base_addr16_23 resb 0 ;low byte of high word of base address
.flags resb 0 ;segment type and misc. flags
.access resb 0 ;highest nibble of segment length and access flags
.base_addr24_31 resb 0 ;highest byte of base address
endstruc
;interrupt descriptor
struc interrupt_descriptor
.offset0_15 resw 0 ;low word of handler offset
.selector0_15 resw 0 ;segment selector
.zero_byte resb 0 ;unused
.flags resb 0 ;flags
.offset16_31 resw 0 ;high word of handler offset
endstruc
;gdt data
gdt_reg dw gdt_size,0,0
dummy_dscr:
istruc segment_descriptor
resw 0
resw 0
resb 0
resb 0
resb 0
iend
code32_dscr:
istruc segment_descriptor
resw 0ffffh
resw 0
resw 0
resb 9ah
resb 0cfh
resb 0
iend
data32_dscr:
istruc segment_descriptor
resw 0ffffh
resw 0
resw 0
resb 92h
resb 0cfh
resb 0
iend
core32_dscr:
istruc segment_descriptor
resw 0ffffh
resw 0
resw 0
resb 9ah
resb 0
resb 0
iend
data16_dscr:
istruc segment_descriptor
resw 0ffffh
resw 0
resw 0
resb 92h
resb 0
resb 0
iend
gdt_size=-$(dummy_dscr)
;task selector
taskselector:
task_backlink resw 0,0
task_esp0 resd 0
task_ss0 resw 0,0
task_esp1 resd 0
task_ss1 resw 0,0
task_esp2 resd 0
tasl_ss2 resw 0,0
task_cr3 resd 0
task_eip resd 0
task_eflags resd 0
task_eax resd 0
task_ecx resd 0
task_edx resd 0
task_ebx resd 0
task_esp resd 0
task_ebp resd 0
task_esi resd 0
task_edi resd 0
task_es resw 0,0
task_cs resw 0,0
task_ss resw 0,0
task_ds resw 0,0
task_fs resw 0,0
task_gs resw 0,0
task_idt resw 0,0
task_trap resw 0
task_iomapbase resw 0
task_iobitmap times 128 resb 0
endtaskselector1:
systemtaskselector equ $-gdt_reg
systemtaskdescriptor:
resw 103 ;limit/length
resb 0 ;base 0
resb 0 ;base 0
resb 10001001b ;p=1, dpl=00, dt=0, available (1,0,0,0,1001)
resb 01000000b ;g=1, d=1, 0, avl=0, 1111=f: limit/length (1,1,0,0,1111)
resb 0 ;base 0
endsystaskdesc:
lgdt [fword ds:gdt_reg]
cli
mov eax,cr0
or al,1
mov cr0,eax
jmp start32
start32:
mov ax, systemtaskselector
ltr ax
ret
Re:kinit() not assembling under nasm
Posted: Sun Jun 15, 2003 7:26 am
by Pype.Clicker
is the root of your problem, imho ... but as i don't exactly get what the hell you're trying to do with all your structs, etc, i hardly can say what you should write instead.
looks pretty weird too.
$-<label> is supposed to be the offset from the label to the current position.
Re:kinit() not assembling under nasm
Posted: Sun Jun 15, 2003 7:34 am
by unknown user
errors (with $-dummy_dscr instead of -$dummy_dscr):
Code: Select all
init.asm:37: TIMES value -359 is negative
init.asm:46: TIMES value -351 is negative
init.asm:55: TIMES value -152 is negative
init.asm:64: TIMES value -144 is negative
init.asm:65: parser: instruction expected
init.asm:113: parser: expecting ]
Re:kinit() not assembling under nasm
Posted: Sun Jun 15, 2003 9:05 am
by Pype.Clicker
one think you should do is declare "GDTREG" *after* your GDt so that the gdt size can be computed when the GDTREG is assembled (and no longer need 2 passes).
i also think you get "resb" wrong:
while 'db <value>' reserves a byte in the current section and give it the declared <value>, "resb <amount>" reserves <amount> bytes and doesn't give them any value.
thus "resb 0" does nothing but generating problems.
Re:kinit() not assembling under nasm
Posted: Sun Jun 15, 2003 10:04 am
by df
the old nasm (0.97 or 98) used to give me mega problems with istrtuc so much so i stopped using istruc . it was too much hassle.
Re:kinit() not assembling under nasm
Posted: Sun Jun 15, 2003 5:46 pm
by unknown user
Pype.Clicker wrote:
while 'db <value>' reserves a byte in the current section and give it the declared <value>, "resb <amount>" reserves <amount> bytes and doesn't give them any value.
thus "resb 0" does nothing but generating problems.
ah. ok. that's probably the problem ^^;. people on EFnet #asm keep telling me that db doesn't reserve a byte in nasm. shame on them.
Re:kinit() not assembling under nasm
Posted: Sun Jun 15, 2003 6:00 pm
by Tim
The people in Freenode #osdev wouldn't have told you that.
Re:kinit() not assembling under nasm
Posted: Mon Jun 16, 2003 10:08 am
by unknown user
ok. i took out all the istrucs and fixed everything else, but there's one problem. nasm doesn't recognize the metacommand "fword".
Re:kinit() not assembling under nasm
Posted: Mon Jun 16, 2003 12:07 pm
by Pype.Clicker
no, and there's no need for a fword, anyway. If you want to make a lgdt command or something alike, just use lgdt[memory address].
size information in nasm are only required when you use a memory operand and a constant operand, so that nasm know what size your memory operand is.
Re:kinit() not assembling under nasm
Posted: Mon Jun 16, 2003 12:12 pm
by unknown user
i fixed it, but i'm getting even more errors. current code:
Code: Select all
init:
;segment descriptor
struc segment_descriptor
.seg_length0_15 dw 0 ;low word of segment length
.base_addr0_15 dw 0 ;low word of base address
.base_addr16_23 db 0 ;low byte of high word of base address
.flags db 0 ;segment type and misc. flags
.access db 0 ;highest nibble of segment length and access flags
.base_addr24_31 db 0 ;highest byte of base address
endstruc
;interrupt descriptor
struc interrupt_descriptor
.offset0_15 dw 0 ;low word of handler offset
.selector0_15 dw 0 ;segment selector
.zero_byte db 0 ;unused
.flags db 0 ;flags
.offset16_31 dw 0 ;high word of handler offset
endstruc
;gdt data
gdt_reg:
dw 0x00ff
dd gdt_reg
dw 0
dw 0x0000,0x0000,0x0000,0x0000
dw 0xffff,0x0000,0x9b00,0x00ef
dw 0xffff,0x0000,0x9300,0x008f
gdt_reg_end:
;task selector
taskselector:
task_backlink dw 0,0
task_esp0 dd 0
task_ss0 dw 0,0
task_esp1 dd 0
task_ss1 dw 0,0
task_esp2 dd 0
tasl_ss2 dw 0,0
task_cr3 dd 0
task_eip dd 0
task_eflags dd 0
task_eax dd 0
task_ecx dd 0
task_edx dd 0
task_ebx dd 0
task_esp dd 0
task_ebp dd 0
task_esi dd 0
task_edi dd 0
task_es dw 0,0
task_cs dw 0,0
task_ss dw 0,0
task_ds dw 0,0
task_fs dw 0,0
task_gs dw 0,0
task_idt dw 0,0
task_trap dw 0
task_iomapbase dw 0
task_iobitmap times 128 db 0
endtaskselector1:
systemtaskselector equ $-taskselector
systemtaskdescriptor:
dw 103 ;limit/length
db 0 ;base 0
db 0 ;base 0
db 10001001b ;p=1, dpl=00, dt=0, available (1,0,0,0,1001)
db 01000000b ;g=1, d=1, 0, avl=0, 1111=f: limit/length (1,1,0,0,1111)
db 0 ;base 0
endsystaskdesc:
lgdt [gdt_reg]
cli
mov eax,cr0
or al,1
mov cr0,eax
jmp start32
start32:
mov ax, [systemtaskselector]
ltr ax
ret
Re:kinit() not assembling under nasm
Posted: Mon Jun 16, 2003 1:08 pm
by unknown user
got it to assemble. thanks, guys ^^;.
Re:kinit() not assembling under nasm
Posted: Tue Jun 17, 2003 1:51 am
by Pype.Clicker
if you have a few sheet of paper to use, i would suggest you to print the NASM manual and go through it if you plan to use assembler for a bit more than just the bootloader. Nasm's preprocessor is *very* powerful (much more than GCC's one, for instance)