Fixed MOVSX,MOVZX, and passed initial paging test
This commit is contained in:
Jeff Parsons
jeffpar
parent
df838b8f86
commit
f3646d206f
10 changed files with 403 additions and 128 deletions
|
|
@ -1,7 +1,7 @@
|
|||
all: tests.json
|
||||
|
||||
tests.com: tests.nasm
|
||||
nasm -f bin tests.nasm -l tests.lst -o tests.com
|
||||
tests.com: tests.nasm ../inc/dos.inc ../inc/misc.inc ../inc/x86.inc
|
||||
nasm -i../inc/ -f bin tests.nasm -l tests.lst -o tests.com
|
||||
|
||||
tests.json: tests.com
|
||||
node ../../../modules/filedump/bin/filedump --file=tests.com --output=tests.json --overwrite
|
||||
|
|
|
|||
|
|
@ -4,48 +4,41 @@
|
|||
;
|
||||
; When used as a ROM, it should be installed at physical address 983296 (0xf0100) and aliased at
|
||||
; physical address 4294902016 (0xffff0100). The jump at jmpStart should align with the CPU reset
|
||||
; address (%0xfffffff0), which will transfer control to 0xf000:0x0100.
|
||||
; address (%0xfffffff0), which will transfer control to 0xf000:0x0100. From that point on,
|
||||
; all memory accesses should remain within the first 1Mb.
|
||||
;
|
||||
; The code which attempts to update myGDT and addrGDT will have no effect when installed as a ROM,
|
||||
; which is fine, because those data structures are predefined with appropriate ROM-based addresses.
|
||||
;
|
||||
; See the machine definition file in /modules/pcjs/bin/romtests.json for a sample ROM configuration.
|
||||
; See the machine definition file in /modules/pcjs/bin/romtests.json for a configuration that can
|
||||
; load this file as a ROM image.
|
||||
;
|
||||
; PMODE_32BIT Notes
|
||||
; PROT32 Notes
|
||||
; -----------------
|
||||
; PMODE_32BIT is NOT enabled by default, because based on what I've seen in VirtualBox (and notes
|
||||
; PROT32 is NOT enabled by default, because based on what I've seen in VirtualBox (and notes
|
||||
; at http://geezer.osdevbrasil.net/johnfine/segments.htm), if CS is loaded with a 32-bit code segment
|
||||
; while in protected-mode and we then return to real-mode, even if we immediately perform a FAR JMP
|
||||
; with a real-mode CS, the base and limit of CS will be updated, but other CS attributes, like EXT_BIG,
|
||||
; will NOT be updated. As a result, the processor will crash as soon as it starts executing 16-bit
|
||||
; real-mode code, because it's being misinterpreted as 32-bit code, and there doesn't appear to be
|
||||
; anything you can do about it from real-mode.
|
||||
; while in protected-mode and we then return to real-mode, even if we immediately perform a FAR jump
|
||||
; with a real-mode CS, the base of CS will be updated, but all the other segment attributes, like
|
||||
; EXT_BIG, remain unchanged. As a result, the processor will crash as soon as it starts executing
|
||||
; 16-bit real-mode code, because it's being misinterpreted as 32-bit code, and there doesn't appear
|
||||
; to be anything you can do about it from real-mode.
|
||||
;
|
||||
; The work-around: switch to a 16-bit code segment BEFORE returning to real-mode.
|
||||
; The work-around: load CS with a 16-bit code segment BEFORE returning to real-mode.
|
||||
;
|
||||
; "Unreal mode" works by setting other segment registers (eg, DS, ES) to 32-bit segments, not CS.
|
||||
; SS should not be set to a 32-bit segment either, because that causes implicit pushes to use ESP
|
||||
; instead of SP, even in real-mode.
|
||||
;
|
||||
; TODO: Verify that a 80386 cannot successfully return to real-mode when CS contains a 32-bit code segment.
|
||||
;
|
||||
cpu 386
|
||||
org 0x100
|
||||
section .text
|
||||
|
||||
%include "dos.inc"
|
||||
%include "misc.inc"
|
||||
%include "x86.inc"
|
||||
|
||||
bits 16
|
||||
|
||||
ACC_TYPE_SEG equ 0x1000
|
||||
ACC_PRESENT equ 0x8000
|
||||
ACC_TYPE_CODE equ 0x0800
|
||||
ACC_TYPE_READABLE equ 0x0200
|
||||
ACC_TYPE_WRITABLE equ 0x0200
|
||||
ACC_TYPE_CODE_READABLE equ 0x1a00
|
||||
ACC_TYPE_DATA_WRITABLE equ 0x1200
|
||||
|
||||
EXT_NONE equ 0x0000
|
||||
EXT_BIG equ 0x0040
|
||||
|
||||
;
|
||||
; If we built our data structures in RAM, we might use the first page of RAM (0x0000-0x0fff) like so:
|
||||
;
|
||||
|
|
@ -66,25 +59,9 @@ EXT_BIG equ 0x0040
|
|||
;RAM_RETF equ 0x0d10
|
||||
|
||||
CSEG_REAL equ 0xf000
|
||||
CSEG_PROT equ 0x0008
|
||||
DSEG_PROT equ 0x0010
|
||||
|
||||
CR0_MSW_PE equ 0x0001
|
||||
|
||||
;
|
||||
; The "set" macro initializes a register to the specified value (eg, "set eax,0")
|
||||
;
|
||||
%macro set 2
|
||||
%ifnum %2
|
||||
%if %2 = 0
|
||||
xor %1,%1
|
||||
%else
|
||||
mov %1,%2
|
||||
%endif
|
||||
%else
|
||||
mov %1,%2
|
||||
%endif
|
||||
%endmacro
|
||||
CSEG_PROT16 equ 0x0008
|
||||
CSEG_PROT32 equ 0x0010
|
||||
DSEG_PROT16 equ 0x0018
|
||||
|
||||
;
|
||||
; The "defDesc" macro defines a descriptor, given a name (%1), base (%2), limit (%3), type (%4), and ext (%5)
|
||||
|
|
@ -118,14 +95,29 @@ CR0_MSW_PE equ 0x0001
|
|||
%endmacro
|
||||
|
||||
start: cli
|
||||
;
|
||||
; Test unsigned 32-bit multiplication and division
|
||||
;
|
||||
mov eax,0x44332211
|
||||
mov ebx,eax
|
||||
mov ecx,0x88776655
|
||||
mul ecx
|
||||
div ecx
|
||||
cmp eax,ebx
|
||||
je near initGDT ; apparently we have to tell NASM "near" because this is a forward reference
|
||||
jne near exitErr ; apparently we have to tell NASM "near" because this is a forward reference
|
||||
|
||||
xor dx,dx
|
||||
mov ds,dx ; DS -> 0x0000
|
||||
;
|
||||
; Test moving a segment register to a 32-bit register
|
||||
;
|
||||
mov eax,ds
|
||||
test eax,eax
|
||||
jnz near exitErr
|
||||
|
||||
jmp initGDT
|
||||
times 32768 nop ; lots of NOPs to force a 16-bit conditional jump
|
||||
|
||||
;
|
||||
; storeDesc(EBX=base, ECX=limit, DX=type, AX=ext, DI=address of descriptor)
|
||||
;
|
||||
|
|
@ -150,19 +142,15 @@ storeDesc:
|
|||
ret
|
||||
|
||||
addrGDT:dw myGDTEnd - myGDT - 1 ; 16-bit limit of myGDT
|
||||
dw myGDT, 0xffff ; 32-bit base address of myGDT (works as long as we're aliased at 0xffff0000)
|
||||
dw myGDT, 0x000f ; 32-bit base address of myGDT
|
||||
|
||||
;
|
||||
; TODO: Why do I need to provide a 2nd parameter for "defDesc NULL"? Is this a NASM 0.98.x bug?
|
||||
;
|
||||
myGDT: defDesc NULL,0 ; the first descriptor in any descriptor table is always a dud (it corresponds to the null selector)
|
||||
%ifdef PMODE_32BIT
|
||||
defDesc CSEG_PROT,0x000f0000,0x0000ffff,ACC_TYPE_CODE_READABLE,EXT_BIG
|
||||
defDesc DSEG_PROT,0x00000000,0x000fffff,ACC_TYPE_DATA_WRITABLE,EXT_BIG
|
||||
%else
|
||||
defDesc CSEG_PROT,0x000f0000,0x0000ffff,ACC_TYPE_CODE_READABLE,EXT_NONE
|
||||
defDesc DSEG_PROT,0x00000000,0x000fffff,ACC_TYPE_DATA_WRITABLE,EXT_NONE
|
||||
%endif
|
||||
defDesc CSEG_PROT16,0x000f0000,0x0000ffff,ACC_TYPE_CODE_READABLE,EXT_NONE
|
||||
defDesc CSEG_PROT32,0x000f0000,0x0000ffff,ACC_TYPE_CODE_READABLE,EXT_BIG
|
||||
defDesc DSEG_PROT16,0x00000000,0x000fffff,ACC_TYPE_DATA_WRITABLE,EXT_NONE
|
||||
myGDTEnd:
|
||||
|
||||
initGDT:
|
||||
|
|
@ -173,8 +161,9 @@ initGDT:
|
|||
xor eax,eax
|
||||
mov ax,cs
|
||||
shl eax,4
|
||||
setDesc CSEG_PROT,eax,0x0000ffff,ACC_TYPE_CODE_READABLE,EXT_NONE
|
||||
setDesc DSEG_PROT,0x0,0x000fffff,ACC_TYPE_DATA_WRITABLE,EXT_NONE
|
||||
setDesc CSEG_PROT16,eax,0x0000ffff,ACC_TYPE_CODE_READABLE,EXT_NONE
|
||||
setDesc CSEG_PROT32,eax,0x0000ffff,ACC_TYPE_CODE_READABLE,EXT_BIG
|
||||
setDesc DSEG_PROT16,0x0,0x000fffff,ACC_TYPE_DATA_WRITABLE,EXT_NONE
|
||||
sub edi,RAM_GDT
|
||||
dec edi
|
||||
mov [RAM_GDTR],di
|
||||
|
|
@ -182,65 +171,177 @@ initGDT:
|
|||
mov word [RAM_RETF+2],cs
|
||||
%else
|
||||
;
|
||||
; This code will have no effect if we're in ROM (but in that case, everything should already be initialized correctly)
|
||||
; This code fixes the GDT and all our FAR jumps if we're running in RAM
|
||||
;
|
||||
xor eax,eax
|
||||
mov ax,cs
|
||||
shl eax,4 ; EAX = base address of the current CS
|
||||
shl eax,4 ; EAX == base address of the current CS
|
||||
mov edx,eax ; save it in EDX
|
||||
mov [cs:myGDT+CSEG_PROT+2],ax ; update the base portions of the descriptor for CSEG_PROT
|
||||
mov [cs:myGDT+CSEG_PROT16+2],ax ; update the base portions of the descriptor for CSEG_PROT16 and CSEG_PROT32
|
||||
mov [cs:myGDT+CSEG_PROT32+2],ax
|
||||
shr eax,16
|
||||
mov [cs:myGDT+CSEG_PROT+4],al
|
||||
mov [cs:myGDT+CSEG_PROT+7],ah
|
||||
mov [cs:myGDT+CSEG_PROT16+4],al
|
||||
mov [cs:myGDT+CSEG_PROT32+4],al
|
||||
mov [cs:myGDT+CSEG_PROT16+7],ah
|
||||
mov [cs:myGDT+CSEG_PROT32+7],ah
|
||||
mov eax,edx ; recover the base address of the current CS
|
||||
add eax,myGDT ; EAX = physical address of myGDT
|
||||
add eax,myGDT ; EAX == physical address of myGDT
|
||||
mov [cs:addrGDT+2],eax ; update the 32-bit base address of myGDT in addrGDT
|
||||
%ifdef PMODE_32BIT
|
||||
mov [cs:jmpReal+5],cs ; update the segment of the far jmp that returns us to real-mode
|
||||
%else
|
||||
mov [cs:jmpReal+3],cs
|
||||
%endif
|
||||
mov [cs:jmpStart+3],cs ; ditto for the far jmp that returns us to the start of the image
|
||||
mov ax,cs
|
||||
%ifdef PROT32
|
||||
mov [cs:jmpReal+5],ax ; update the segment of the FAR jump that returns us to real-mode
|
||||
%else
|
||||
mov [cs:jmpReal+3],ax
|
||||
%endif
|
||||
mov [cs:jmpStart+3],ax ; ditto for the FAR jump that returns us to the start of the image
|
||||
%endif
|
||||
;
|
||||
; Now we want to build a page directory and a page table, but we need two pages of
|
||||
; 4K-aligned physical memory. We can use hard-coded addresses in we're running in ROM,
|
||||
; otherwise we ask DOS for some memory.
|
||||
;
|
||||
cmp ax,CSEG_REAL
|
||||
mov edi,0x1000 ; default to the 2nd physical page in low memory
|
||||
je initPages
|
||||
mov bx,0x1000 ; 4K paragraphs == 64K bytes
|
||||
mov ah,DOS_SETBLOCK ; resize the current block so we can allocate a new block
|
||||
int INT_DOS
|
||||
jnc allocPages
|
||||
exitErrDOSMem:
|
||||
mov dx,errDOSMem
|
||||
exitErrDOS:
|
||||
mov ah,DOS_STD_CON_STRING_OUTPUT
|
||||
int INT_DOS
|
||||
int INT_DOSEXIT
|
||||
|
||||
errDOSMem:
|
||||
db "Insufficient memory",CR,LF,'$'
|
||||
|
||||
exitErr:nop
|
||||
int3
|
||||
jmp exitErr
|
||||
|
||||
allocPages:
|
||||
mov bx,0x1000 ; 4K paragraphs == 64K bytes
|
||||
mov ah,DOS_ALLOC
|
||||
int INT_DOS
|
||||
jc errDOSMem
|
||||
;
|
||||
; AX == segment of 64K memory block
|
||||
;
|
||||
movzx eax,ax
|
||||
shl eax,4
|
||||
add eax,0xfff
|
||||
and eax,~0xfff
|
||||
mov edi,eax ; EDI == first physical 4K-aligned page within the 64K
|
||||
|
||||
initPages:
|
||||
mov esi,edi ; ESI == saved copy of EDI
|
||||
;
|
||||
; Build a page directory at EDI with only 1 valid PDE (the first one)
|
||||
;
|
||||
cld
|
||||
mov eax,edi
|
||||
add eax,0x1000 ; EAX == page frame address (of the next page)
|
||||
or eax,PTE_USER | PTE_READWRITE | PTE_PRESENT
|
||||
stosd
|
||||
mov ecx,1024-1 ; ECX == number of (remaining) PDEs to write
|
||||
sub eax,eax
|
||||
rep stosd
|
||||
;
|
||||
; Build a page table at EDI with only 256 (out of 1024) valid PTEs, which will map the first 1Mb of the
|
||||
; first 4Mb as linear == physical.
|
||||
;
|
||||
mov eax,PTE_USER | PTE_READWRITE | PTE_PRESENT
|
||||
mov ecx,256 ; ECX == number of PTEs to write
|
||||
initPT: stosd
|
||||
add eax,0x1000
|
||||
loop initPT
|
||||
mov ecx,1024-256 ; ECX == number of (remaining) PTEs to write
|
||||
sub eax,eax
|
||||
rep stosd
|
||||
|
||||
goProt: o32 lgdt [cs:addrGDT]
|
||||
mov cr3,esi
|
||||
mov eax,cr0
|
||||
%ifdef PAGING
|
||||
or eax,CR0_MSW_PE | CR0_PG
|
||||
%else
|
||||
or eax,CR0_MSW_PE
|
||||
mov cr0,eax
|
||||
nop
|
||||
jmpProt:
|
||||
jmp CSEG_PROT:toProt
|
||||
|
||||
toProt:
|
||||
%ifdef PMODE_32BIT
|
||||
bits 32 ; only if we define the CSEG_PROT descriptor with EXT_BIG
|
||||
%endif
|
||||
mov ax,DSEG_PROT
|
||||
mov cr0,eax
|
||||
jmpProt:
|
||||
jmp CSEG_PROT32:toProt32
|
||||
|
||||
toProt32:
|
||||
bits 32
|
||||
|
||||
mov ax,DSEG_PROT16
|
||||
mov ds,ax
|
||||
mov es,ax
|
||||
;
|
||||
; Do some protected-mode tests now...
|
||||
;
|
||||
;
|
||||
; Test moving a segment register to a 32-bit memory location
|
||||
;
|
||||
mov edx,[0x0000] ; save the DWORD at 0x0000:0x0000 in EDX
|
||||
or eax,-1
|
||||
mov [0x0000],eax
|
||||
mov [0x0000],ds
|
||||
mov eax,ds
|
||||
cmp eax,[0x0000]
|
||||
err1: jne err1
|
||||
mov [0x0000],edx ; restore the DWORD at 0x0000:0x0000 from EDX
|
||||
|
||||
;
|
||||
; Test moving a byte to a 32-bit register with sign-extension
|
||||
;
|
||||
movsx eax,byte [0xfffff]
|
||||
cmp eax,0xffffff80
|
||||
err2: jne err2
|
||||
|
||||
;
|
||||
; Test moving a word to a 32-bit register with sign-extension
|
||||
;
|
||||
movsx eax,word [0xffffe]
|
||||
cmp eax,0xffff80fc
|
||||
err3: jne err3
|
||||
|
||||
;
|
||||
; Test moving a byte to a 32-bit register with zero-extension
|
||||
;
|
||||
movzx eax,byte [0xfffff]
|
||||
cmp eax,0x00000080
|
||||
err4: jne err4
|
||||
|
||||
;
|
||||
; Test moving a word to a 32-bit register with zero-extension
|
||||
;
|
||||
movzx eax,word [0xffffe]
|
||||
cmp eax,0x000080fc
|
||||
err5: jne err5
|
||||
|
||||
;
|
||||
; Return to real-mode now, after first loading CS with a 16-bit code segment
|
||||
;
|
||||
jmp CSEG_PROT16:toProt16
|
||||
|
||||
toProt16:
|
||||
bits 16
|
||||
|
||||
goReal: mov eax,cr0
|
||||
and eax,~CR0_MSW_PE
|
||||
and eax,~(CR0_MSW_PE | CR0_PG)
|
||||
mov cr0,eax
|
||||
nop
|
||||
jmpReal:
|
||||
jmp CSEG_REAL:toReal
|
||||
|
||||
toReal:
|
||||
%ifdef PMODE_32BIT
|
||||
bits 16 ; only if we define the CSEG_PROT descriptor with EXT_BIG
|
||||
%endif
|
||||
mov ax,cs
|
||||
cmp ax,CSEG_REAL ; is CS equal to 0xf000?
|
||||
je near jmpStart ; yes
|
||||
int 0x20 ; no, so assume we're running under DOS and exit
|
||||
|
||||
;
|
||||
; Fill the remaining space with NOPs until we get to target offset 0xFFF0.
|
||||
; Note that we subtract 0x100 from the target offset because we're ORG'ed at 0x100.
|
||||
;
|
||||
je near jmpStart ; yes, so loop around, only because we have nowhere else to go
|
||||
int INT_DOSEXIT ; no, so assume we're running under DOS and exit
|
||||
;
|
||||
; Fill the remaining space with NOPs until we get to target offset 0xFFF0.
|
||||
; Note that we subtract 0x100 from the target offset because we're ORG'ed at 0x100.
|
||||
;
|
||||
times 0xfff0-0x100-($-$$) nop
|
||||
|
||||
jmpStart:
|
||||
|
|
@ -249,4 +350,4 @@ jmpStart:
|
|||
db 0x20
|
||||
db '04/04/15'
|
||||
db 0xFC ; 0000FFFE FC (Model ID byte)
|
||||
db 0x00 ; 0000FFFF 00 (location of checksum byte)
|
||||
db 0x80 ; 0000FFFF 80 (normally, location of a checksum byte)
|
||||
|
|
|
|||
Loading…
Reference in a new issue