; App written by randall ported to KolibriOS and MenuetOS64 by macgub (www.macgub.hekko.pl).
; Now it use static memory, it is mixed 32bit code and SSE2 instructions.
include '../../../macros.inc'
use32
org 0x0
db 'MENUET01' ; 8 byte id
dd 0x01 ; header version
dd START ; start of code
dd IMG_END ; size of image
dd I_END ;0x100000 ; memory for app
dd I_END ;0xbffff ; esp
dd 0x0 , 0x0 ; I_Param , I_Icon
START: ; start of execution
call draw_window
call Main
call draw_from_buffer
still:
; mov eax,23 ; wait here for event
; mov ebx,timeout
; int 0x40
; mov eax,11 ; check for event no wait
mov eax,10 ; wait for event
int 0x40
cmp eax,1 ; redraw request ?
je red
cmp eax,2 ; key in buffer ?
je key
cmp eax,3 ; button in buffer ?
je button
jmp noclose
red: ; redraw
call draw_window
call draw_from_buffer
jmp still
key: ; key
mov eax,2 ; just read it and ignore
int 0x40
shr eax,8
cmp eax, 27
jne still
mov eax, -1
int 0x40
button: ; button
mov eax,17 ; get id
int 0x40
cmp ah,1 ; button id=1 ?
jne noclose
mov eax,-1 ; close this program
int 0x40
noclose:
jmp still
draw_from_buffer:
mov eax,7
mov ebx,screen4
mov ecx,IMG_WIDTH*65536+IMG_HEIGHT
mov edx,0*65536+0
int 0x40
ret
;-------------------------------------------------------------------------------
; NAME: shuf
;-------------------------------------------------------------------------------
macro shuf d*,s*,z*,y*,x*,w* {
shufps d,s,(z shl 6) or (y shl 4) or (x shl 2) or w
}
;-------------------------------------------------------------------------------
; NAME: length3
; IN: xmm0.xyz input vector
; OUT: xmm0.xyzw vector length
;-------------------------------------------------------------------------------
macro length3 {
mulps xmm0,xmm0
movaps xmm1,xmm0
movaps xmm2,xmm0
shufps xmm0,xmm0,0x00
shufps xmm1,xmm1,0x55
shufps xmm2,xmm2,0xaa
addps xmm0,xmm1
addps xmm0,xmm2
sqrtps xmm0,xmm0
}
;-------------------------------------------------------------------------------
; NAME: length4
; IN: xmm0.xyzw input vector
; OUT: xmm0.xyzw vector length
;-------------------------------------------------------------------------------
macro length4 {
mulps xmm0,xmm0
movaps xmm1,xmm0
movaps xmm2,xmm0
movaps xmm3,xmm0
shufps xmm0,xmm0,0x00
shufps xmm1,xmm1,0x55
shufps xmm2,xmm2,0xaa
shufps xmm3,xmm3,0xff
addps xmm0,xmm1
addps xmm2,xmm3
addps xmm0,xmm2
sqrtps xmm0,xmm0
}
;-------------------------------------------------------------------------------
; NAME: normalize3
; IN: xmm0.xyz input vector
; OUT: xmm0.xyz normalized vector
;-------------------------------------------------------------------------------
macro normalize3 {
movaps xmm3,xmm0
mulps xmm0,xmm0
movaps xmm1,xmm0
movaps xmm2,xmm0
shufps xmm0,xmm0,0x00
shufps xmm1,xmm1,0x55
shufps xmm2,xmm2,0xaa
addps xmm0,xmm1
addps xmm0,xmm2
rsqrtps xmm0,xmm0
mulps xmm0,xmm3
}
;-------------------------------------------------------------------------------
; NAME: cross
; IN: xmm0.xyz first input vector
; IN: xmm1.xyz second input vector
; OUT: xmm0.xyz cross product result
;-------------------------------------------------------------------------------
macro cross {
movaps xmm2,xmm0
movaps xmm3,xmm1
shuf xmm0,xmm0,3,0,2,1
shuf xmm1,xmm1,3,1,0,2
shuf xmm2,xmm2,3,1,0,2
shuf xmm3,xmm3,3,0,2,1
mulps xmm0,xmm1
mulps xmm2,xmm3
subps xmm0,xmm2
}
;-------------------------------------------------------------------------------
; NAME: dot3
; IN: xmm0.xyz first input vector
; IN: xmm1.xyz second input vector
; OUT: xmm0.xyzw dot product result
;-------------------------------------------------------------------------------
macro dot3 {
mulps xmm0,xmm1
movaps xmm1,xmm0
movaps xmm2,xmm0
shufps xmm0,xmm0,0x00
shufps xmm1,xmm1,0x55
shufps xmm2,xmm2,0xaa
addps xmm0,xmm1
addps xmm0,xmm2
}
;-------------------------------------------------------------------------------
; NAME: dot4
; IN: xmm0.xyzw first input vector
; IN: xmm1.xyzw second input vector
; OUT: xmm0.xyzw dot product result
;-------------------------------------------------------------------------------
macro dot4 {
mulps xmm0,xmm1
movaps xmm1,xmm0
movaps xmm2,xmm0
movaps xmm3,xmm0
shufps xmm0,xmm0,0x00
shufps xmm1,xmm1,0x55
shufps xmm2,xmm2,0xaa
shufps xmm3,xmm3,0xff
addps xmm0,xmm1
addps xmm2,xmm3
addps xmm0,xmm2
}
;-------------------------------------------------------------------------------
; NAME: qsq
; IN: xmm0.xyzw quaternion
; OUT: xmm0.xyzw squared quaternion
;-------------------------------------------------------------------------------
macro qsq {
movaps xmm3,xmm0
movaps xmm4,xmm0
movaps xmm1,xmm0
dot3
shufps xmm3,xmm3,0xff
movaps xmm5,xmm3
mulps xmm3,xmm3
subps xmm3,xmm0 ; r.___w
mulps xmm4,xmm5
addps xmm4,xmm4 ; r.xyz_
andps xmm3,dqword [g_ClearXYZ]
andps xmm4,dqword [g_ClearW]
orps xmm3,xmm4
movaps xmm0,xmm3
}
;-------------------------------------------------------------------------------
; NAME: qmul
; IN: xmm0.xyzw first quaternion
; IN: xmm1.xyzw second quaternion
; OUT: xmm0.xyzw first quaternion multiplied by second quaternion
;-------------------------------------------------------------------------------
macro qmul {
movaps xmm6,xmm0 ; q1.xyzw
movaps xmm7,xmm1 ; q2.xyzw
dot3
movaps xmm4,xmm6
movaps xmm5,xmm7
shufps xmm4,xmm4,0xff ; q1.wwww
shufps xmm5,xmm5,0xff ; q2.wwww
movaps xmm1,xmm4
mulps xmm1,xmm5
subps xmm1,xmm0
andps xmm1,dqword [g_ClearXYZ]
movaps [xxmm8],xmm1 ; r.000w
movaps xmm0,xmm6
movaps xmm1,xmm7
cross
mulps xmm7,xmm4
mulps xmm6,xmm5
addps xmm6,xmm7
addps xmm0,xmm6
andps xmm0,dqword [g_ClearW]
orps xmm0,[xxmm8]
}
;-------------------------------------------------------------------------------
; NAME: logss
; IN: xmm0.x function argument
; OUT: xmm0.x function result
;-------------------------------------------------------------------------------
macro logss {
maxss xmm0,[g_MinNormPos]
movss xmm1,[g_1_0]
movd edx,xmm0
andps xmm0,dqword [g_InvMantMask]
orps xmm0,xmm1
movaps xmm4,xmm0
subss xmm0,xmm1
addss xmm4,xmm1
shr edx,23
rcpss xmm4,xmm4
mulss xmm0,xmm4
addss xmm0,xmm0
movaps xmm2,xmm0
mulss xmm0,xmm0
sub edx,0x7f
movss xmm4,[g_log_p0]
movss xmm6,[g_log_q0]
mulss xmm4,xmm0
movss xmm5,[g_log_p1]
mulss xmm6,xmm0
movss xmm7,[g_log_q1]
addss xmm4,xmm5
addss xmm6,xmm7
movss xmm5,[g_log_p2]
mulss xmm4,xmm0
movss xmm7,[g_log_q2]
mulss xmm6,xmm0
addss xmm4,xmm5
movss xmm5,[g_log_c0]
addss xmm6,xmm7
cvtsi2ss xmm1,edx
mulss xmm0,xmm4
rcpss xmm6,xmm6
mulss xmm0,xmm6
mulss xmm0,xmm2
mulss xmm1,xmm5
addss xmm0,xmm2
addss xmm0,xmm1
}
;-------------------------------------------------------------------------------
; NAME: QJuliaDist
; IN: xmm0.xyz position
; OUT: xmm0.xyzw distance to the nearest point in quaternion julia set
;-------------------------------------------------------------------------------
align 64
QJuliaDist:
Z equ ebp-16
Zp equ ebp-32
NormZ equ ebp-36
NormZp equ ebp-40
push ebp
mov ebp,esp
sub esp,64
; init Z and Zp
andps xmm0,dqword [g_ClearW]
shuf xmm0,xmm0,0,3,2,1
movups [Z],xmm0
movaps xmm1,dqword [g_UnitW]
movups [Zp],xmm1
; iterate
mov ecx,10
.Iterate:
; compute and update Zp
movups xmm0,[Z]
movups xmm1,[Zp]
qmul
addps xmm0,xmm0
movups [Zp],xmm0
; compute and update Z
movups xmm0,[Z]
qsq
addps xmm0,dqword [g_Quat]
movups [Z],xmm0
; check if squared length of Z is greater than g_EscapeThreshold,
; break the loop if it is
movaps xmm1,xmm0
dot4
movss xmm1,[g_EscapeThreshold]
cmpltss xmm1,xmm0
movd eax,xmm1
cmp eax,0xffffffff
je .IterateEnd
; continue the loop
sub ecx,1
test ecx,ecx
jnz .Iterate
.IterateEnd:
movups xmm0,[Z]
length4
movss [NormZ],xmm0
movups xmm0,[Zp]
length4
movss [NormZp],xmm0
movss xmm0,[NormZ]
logss
divss xmm0,[NormZp]
mulss xmm0,[NormZ]
mulss xmm0,[g_0_5]
shufps xmm0,xmm0,0x00
mov esp,ebp
pop ebp
; restore Z,Zp,NormZ,NormZp
ret
;-------------------------------------------------------------------------------
; NAME: Map
; IN: xmm0.xyz position
; OUT: xmm0.xyzw distance to the nearest object from input position
; OUT: eax material ID of the nearest object
;-------------------------------------------------------------------------------
align 64
Map:
; P equ ebp-16
; MinDist equ ebp-32
MatID equ ebp-40
push ebp
mov ebp,esp
sub esp,128
movaps [.P],xmm0
movaps xmm0,dqword [g_255_0]
movaps [.MinDist],xmm0
mov dword [MatID],0
; QJulia
movaps xmm0,[.P]
call QJuliaDist
movaps xmm1,xmm0
cmpltps xmm1,[.MinDist]
movd eax,xmm1
cmp eax,0xffffffff
jne @f
movaps [.MinDist],xmm0
mov dword [MatID],4
@@:
; sphere
movaps xmm0,[.P]
subps xmm0,dqword [g_UnitY]
length3
subps xmm0,dqword [g_1_0]
movaps xmm1,xmm0
cmpltps xmm1,[.MinDist]
movd eax,xmm1
cmp eax,0xffffffff
jne @f
;movaps [MinDist],xmm0
;mov dword [MatID],1
@@:
; plane
movaps xmm0,[.P]
movaps xmm1,dqword [g_UnitY]
dot3
addps xmm0,dqword [g_1_0]
movaps xmm1,xmm0
cmpltps xmm1,[.MinDist]
movd eax,xmm1
cmp eax,0xffffffff
jne @f
movaps [.MinDist],xmm0
mov dword [MatID],2
@@:
; box
xorps xmm0,xmm0
subps xmm0,[.P]
maxps xmm0,[.P]
subps xmm0,dqword [g_BoxSize]
maxps xmm0,dqword [g_0_0]
length3
subps xmm0,dqword [g_BoxEdge]
movaps xmm1,xmm0
cmpltps xmm1,[.MinDist]
movd eax,xmm1
cmp eax,0xffffffff
jne @f
;movaps [MinDist],xmm0
;mov dword [MatID],3
@@:
movaps xmm0,[.MinDist]
mov eax,[MatID]
mov esp,ebp
pop ebp
ret
align 16
.P:
rd 4
.MinDist:
rd 4
; restore P,MinDist,MatID
;-------------------------------------------------------------------------------
; NAME: CastRay
; IN: xmm0.xyz ray origin
; IN: xmm1.xyz ray direction
; OUT: xmm0.xyzw distance from ray orgin to the nearest intersected object
; or -1.0 if there is no intersection
;-------------------------------------------------------------------------------
align 64
CastRay:
; RO equ ebp-16
; RD equ ebp-32
; T equ ebp-48
MatID equ ebp-52
push ebp
mov ebp,esp
sub esp,128
; init stack variables
movaps [.RO],xmm0
movaps [.RD],xmm1
xorps xmm0,xmm0
movaps [.T],xmm0
.March:
; find distance to the nearest object
movaps xmm0,[.RD]
mulps xmm0,[.T]
addps xmm0,[.RO]
call Map
mov [MatID],eax
; return if distance is less than g_HitDist
movaps xmm1,xmm0
cmpltps xmm1,dqword [g_HitDist]
movd eax,xmm1
cmp eax,0xffffffff
je .Hit
; increment T with distance to the nearest object
movaps xmm1,[.T]
addps xmm1,xmm0
movaps [.T],xmm1
; continue loop only if distance is less than g_MaxDist
cmpltps xmm1,dqword [g_MaxDist]
movd eax,xmm1
cmp eax,0xffffffff
je .March
xorps xmm0,xmm0
subps xmm0,dqword [g_1_0]
movaps [.T],xmm0
mov dword [MatID],0
.Hit:
movaps xmm0,[.T]
mov eax,[MatID]
mov esp,ebp
pop ebp
ret
align 16
.RO:
rd 4
.RD:
rd 4
.T:
rd 4
; restore RO,RD,T,MatID
;-------------------------------------------------------------------------------
; NAME: CastShadowRay
; IN: xmm0.xyz ray origin
; IN: xmm1.xyz ray direction
; OUT: xmm0.xyzw visibility factor [0.0, 1.0],
; 0.0 means path is fully blocked,
; 1.0 means path is fully clear
;-------------------------------------------------------------------------------
align 64
CastShadowRay:
; RO equ ebp-16
; RD equ ebp-32
R equ ebp-48
; T equ ebp-64
push ebp
mov ebp,esp
sub esp,128
; init stack variables
movaps [.RO],xmm0
movaps [.RD],xmm1
movaps xmm0,dqword [g_0_01]
movaps [.T],xmm0
movaps xmm0,dqword [g_1_0]
movups [R],xmm0
.March:
; find distance to the nearest object
movaps xmm0,[.RD]
mulps xmm0,[.T]
addps xmm0,[.RO]
call Map
; return 0.0 if distance is less than g_ShadowHitDist
movaps xmm1,xmm0
cmpltps xmm1,dqword [g_ShadowHitDist]
movd eax,xmm1
cmp eax,0xffffffff
je .Hit
; compute R
movaps xmm1,xmm0
rcpps xmm2,[.T]
mulps xmm1,xmm2
mulps xmm1,dqword [g_16_0]
movups xmm2,[R]
minps xmm2,xmm1
movups [R],xmm2
; increment T with distance to the nearest object
movaps xmm1,[.T]
addps xmm1,xmm0
movaps [.T],xmm1
; continue loop only if distance is less than g_ShadowMaxDist
cmpltps xmm1,dqword [g_ShadowMaxDist]
movd eax,xmm1
cmp eax,0xffffffff
je .March
; return (R,R,R,R)
movups xmm0,[R]
mov esp,ebp
pop ebp
ret
.Hit:
; return (0,0,0,0)
xorps xmm0,xmm0
mov esp,ebp
pop ebp
ret
; restore RO,RD,R,T
align 16
.RO:
rd 4
.RD:
rd 4
.T:
rd 4
;-------------------------------------------------------------------------------
; NAME: ComputeNormal
; IN: xmm0.xyz position
; OUT: xmm0.xyz normal vector
;-------------------------------------------------------------------------------
align 64
ComputeNormal:
P equ ebp-16
N equ ebp-32
push ebp
mov ebp,esp
sub esp,128
movups [P],xmm0
; compute x coordinate
addps xmm0,dqword [g_NormalDX]
call Map
movss [N+0],xmm0
movups xmm0,[P]
subps xmm0,dqword [g_NormalDX]
call Map
movss xmm1,[N+0]
subss xmm1,xmm0
movss [N+0],xmm1
; compute y coordinate
movups xmm0,[P]
addps xmm0,dqword [g_NormalDY]
call Map
movss [N+4],xmm0
movups xmm0,[P]
subps xmm0,dqword [g_NormalDY]
call Map
movss xmm1,[N+4]
subss xmm1,xmm0
movss [N+4],xmm1
; compute z coordinate
movups xmm0,[P]
addps xmm0,dqword [g_NormalDZ]
call Map
movss [N+8],xmm0
movups xmm0,[P]
subps xmm0,dqword [g_NormalDZ]
call Map
movss xmm1,[N+8]
subss xmm1,xmm0
movss [N+8],xmm1
; normalize
movups xmm0,[N]
normalize3
mov esp,ebp
pop ebp
ret
; restore P,N
;-------------------------------------------------------------------------------
; NAME: Shade
; IN: xmm0.xyz position
; IN: xmm1.xyz normal vector
; IN: edi material ID
; OUT: xmm0.xyz color
;-------------------------------------------------------------------------------
align 64
Shade:
; P equ ebp-16
; N equ ebp-32
; RGB equ ebp-48
C equ ebp-64
NdotL equ ebp-80
L equ ebp-96
; AOScale equ ebp-112
; AO equ ebp-128
Temp equ ebp-144
Idx equ ebp-148
MatID equ ebp-152
push ebp
mov ebp,esp
sub esp,256
movaps [.P],xmm0
movaps [.N],xmm1
mov [MatID],edi
;
; AO
;
xorps xmm0,xmm0
movaps [.AO],xmm0
movaps xmm0,dqword [g_10_0]
movaps [.AOscale],xmm0
mov dword [Idx],0
.AOLoop:
cvtsi2ss xmm0,[Idx]
shufps xmm0,xmm0,0x00
mulps xmm0,xmm0
mulps xmm0,dqword [g_0_015]
addps xmm0,dqword [g_0_01]
movups [Temp],xmm0
mulps xmm0,[.N]
addps xmm0,[.P]
call Map
movups xmm1,[Temp]
subps xmm1,xmm0
mulps xmm1,[.AOscale]
movaps xmm0,[.AO]
addps xmm0,xmm1
movaps [.AO],xmm0
movaps xmm0,[.AOscale]
mulps xmm0,dqword [g_0_5]
movaps [.AOscale],xmm0
add dword [Idx],1
cmp dword [Idx],5
jne .AOLoop
movaps xmm0,[.AO]
maxps xmm0,dqword [g_0_0]
minps xmm0,dqword [g_1_0]
movaps xmm1,dqword [g_1_0]
subps xmm1,xmm0
movaps [.AO],xmm1
;
; Material ID Switch
;
mov edi,[MatID]
cmp edi,1
je .Mat1
cmp edi,2
je .Mat2
jmp .MatDef
.Mat1:
movaps xmm0,dqword [g_Red]
movaps [.RGB],xmm0
jmp .MatBreak
.Mat2:
movaps xmm0,dqword [g_Green]
movaps [.RGB],xmm0
jmp .MatBreak
.MatDef:
movaps xmm0,dqword [g_1_0]
movaps [.RGB],xmm0
.MatBreak:
;
; Light0 Contribution
;
; compute light vector and "N dot L" value
movaps xmm0,dqword [g_L0Pos]
subps xmm0,[.P]
normalize3
movups [L],xmm0
movups xmm1,[N]
dot3
movups [NdotL],xmm0
; cast shadow ray
movaps xmm0,[.P]
movups xmm1,[L]
call CastShadowRay
movups xmm1,[NdotL]
mulps xmm0,xmm1
mulps xmm0,dqword [g_0_6]
addps xmm0,dqword [g_0_4]
maxps xmm0,dqword [g_0_0]
mulps xmm0,dqword [g_0_7]
mulps xmm0,dqword [.RGB]
movups [C],xmm0
;
; Light1 Contribution
;
; compute light vector and "N dot L" value
movaps xmm0,dqword [g_L1Pos]
subps xmm0,[.P]
normalize3
movups [L],xmm0
movaps xmm1,[.N]
dot3
movups [NdotL],xmm0
; cast shadow ray
movaps xmm0,[.P]
movups xmm1,[L]
call CastShadowRay
movups xmm1,[NdotL]
mulps xmm0,xmm1
mulps xmm0,dqword [g_0_6]
addps xmm0,dqword [g_0_4]
maxps xmm0,dqword [g_0_0]
mulps xmm0,dqword [g_0_3]
mulps xmm0,dqword [.RGB]
movups xmm1,[C]
addps xmm0,xmm1
mulps xmm0,[.AO]
mov esp,ebp
pop ebp
ret
; restore P,N,C,RGB,NdotL,L,AOScale,AO,Temp,Idx,MatID
align 16
.N:
rd 4
.P:
rd 4
.RGB:
rd 4
.AOscale:
rd 4
.AO:
rd 4
;-------------------------------------------------------------------------------
; NAME: ComputeColor
; IN: xmm0.x normalized x coordinate
; IN: xmm0.y normalized y coordinate
; OUT: xmm0.xyz pixel color
;-------------------------------------------------------------------------------
align 64
ComputeColor:
; X equ ebp-32
; Y equ ebp-48
; RD equ ebp-64
P equ ebp-80
N equ ebp-96
MatID equ ebp-100
push ebp
mov ebp,esp
sub esp,128
; save function parameters on the stack
shufps xmm0,xmm0,0x00
shufps xmm1,xmm1,0x00
movaps [.X],xmm0
movaps [.Y],xmm1
; compute z axis
xorps xmm0,xmm0
subps xmm0,dqword [g_CamPos]
normalize3
movaps xmm7,xmm0
; compute x axis
movaps xmm0,dqword [g_UnitY]
movaps xmm1,xmm7
cross
normalize3
movaps xmm6,xmm0
; compute y axis
movaps xmm0,xmm7
movaps xmm1,xmm6
cross
normalize3
; compute ray direction
mulps xmm0,[.Y]
mulps xmm6,[.X]
movaps xmm1,xmm7
mulps xmm1,dqword [g_0_5]
addps xmm7,xmm1
addps xmm0,xmm6
addps xmm0,xmm7
normalize3
movaps [.RD],xmm0
; cast ray
movaps xmm0,dqword [g_CamPos]
movaps xmm1,[.RD]
call CastRay
mov [MatID],eax
; return if there is no intersection
movaps xmm1,xmm0
cmpltps xmm0,dqword [g_0_0]
movd eax,xmm0
cmp eax,0xffffffff
je .Return
; compute intersection point
movaps xmm0,[.RD]
mulps xmm0,xmm1
addps xmm0,dqword [g_CamPos]
movups [P],xmm0
; compute normal vector
call ComputeNormal
movups [N],xmm0
; shade
mov edi,[MatID]
movups xmm0,[P]
movups xmm1,[N]
call Shade
.Return:
mov esp,ebp
pop ebp
ret
; restore X,Y,RD,P,N,MatID
align 16
.X:
rd 4
.Y:
rd 4
.RD:
rd 4
;-------------------------------------------------------------------------------
; NAME: Main
; DESC: Program entry point.
;-------------------------------------------------------------------------------
align 64
Main:
ImgPtr equ ebp-8
Img8Ptr equ ebp-16
X equ ebp-20
Y equ ebp-24
mov ebp,esp
sub esp,128
;
; generate floating point image
;
mov dword [ImgPtr],screen16
mov ebx,screen16
; begin loops
mov dword [Y],0
.LoopY:
mov dword [X],0
.LoopX:
; compute normalized x coordinate [-1.777 , 1.777]
cvtsi2ss xmm0,[X]
divss xmm0,[g_ImgWidth]
subss xmm0,[g_0_5]
addss xmm0,xmm0
mov eax,1.777
movd xmm2,eax
mulss xmm0,xmm2
; compute normalized y coordinate [-1.0 , 1.0]
cvtsi2ss xmm1,[Y]
divss xmm1,[g_ImgHeight]
subss xmm1,[g_0_5]
addss xmm1,xmm1
; compute and write pixel color to the buffer
call ComputeColor
movaps [ebx],xmm0
; advance pixel pointer
add ebx,16
; continue .LoopX
add dword [X],1
cmp dword [X],IMG_WIDTH
jne .LoopX
; continue .LoopY
add dword [Y],1
cmp dword [Y],IMG_HEIGHT
jne .LoopY
;
; convert image to 4 x 8 bpp
;
mov dword [Img8Ptr],screen4
; set dst and src pointers and loop counter
mov edi,screen4
mov esi,[ImgPtr]
mov ecx,IMG_WIDTH*IMG_HEIGHT
@@:
; clamp to [0.0 ,1.0]
movaps xmm0,[esi]
maxps xmm0,dqword [g_0_0]
minps xmm0,dqword [g_1_0]
; convert from [0.0 ,1.0] to [0 ,255]
mulps xmm0,dqword [g_255_0]
cvttps2dq xmm0,xmm0
; pshufb xmm0,dqword [g_ImageConvMask]
; set alpha to 0xff
movd [edi],xmm0
or dword [edi],0xff000000
; continue loop
add esi,16
add edi,4
sub ecx,1
jnz @b
;convert to 24 bit per pixel
mov esi,screen4
mov edi,esi
mov ecx,IMG_WIDTH*IMG_HEIGHT
cld
@@:
lodsd
mov ah,al
push ax
ror eax,16
pop ax
stosd
dec edi
loop @b
mov esp,ebp
ret
; *********************************************
; ******* WINDOW DEFINITIONS AND DRAW ********
; *********************************************
draw_window:
mcall 12, 1 ; function 12:tell os about windowdraw
mcall 48, 4 ;get skin width
lea ecx, [50*65536+IMG_HEIGHT+4+eax] ; [y start] *65536 + [y size] + [skin_height]
mcall 0,<50,IMG_WIDTH+9>,,0x74000000,,labelt ;draw window
mcall 12, 2 ; function 12:tell os about windowdraw
ret
;-------------------------------------------------------------------------------
labelt:
db 'qjulia',0
labelen:
align 4
IMG_WIDTH = 800
IMG_HEIGHT = 600
g_ImgWidth dd 800.0
g_ImgHeight dd 600.0
g_EscapeThreshold dd 16.0
g_MinNormPos dd 0x00800000
g_log_p0 dd -0.789580278884799154124
g_log_p1 dd 16.3866645699558079767
g_log_p2 dd -64.1409952958715622951
g_log_q0 dd -35.6722798256324312549
g_log_q1 dd 312.093766372244180303
g_log_q2 dd -769.691943550460008604
g_log_c0 dd 0.693147180559945
align 16
;g_ImageConvMask db 8,4,0,12,12 dup 0x80
g_InvMantMask dd 4 dup (not 0x7f800000)
g_0_0 dd 4 dup 0.0
g_0_5 dd 4 dup 0.5
g_1_0 dd 4 dup 1.0
g_255_0 dd 4 dup 255.0
g_16_0 dd 4 dup 128.0
g_10_0 dd 4 dup 10.0
g_0_6 dd 4 dup 0.6
g_0_4 dd 4 dup 0.4
g_0_7 dd 4 dup 0.7
g_0_3 dd 4 dup 0.3
g_0_01 dd 4 dup 0.01
g_0_015 dd 4 dup 0.015
g_CamPos dd 1.2,1.4,1.2,1.0
g_UnitY dd 0.0,1.0,0.0,0.0
g_UnitW dd 0.0,0.0,0.0,1.0
g_HitDist dd 4 dup 0.001
g_ShadowHitDist dd 4 dup 0.0005
g_ShadowMaxDist dd 4 dup 10.0
g_MaxDist dd 4 dup 40.0
g_NormalDX dd 0.001,0.0,0.0,0.0
g_NormalDY dd 0.0,0.001,0.0,0.0
g_NormalDZ dd 0.0,0.0,0.001,0.0
g_L0Pos dd 10.0,8.0,-6.0,1.0
g_L1Pos dd -12.0,19.0,6.0,1.0
g_Red dd 1.0,1.0,1.0,1.0
g_Green dd 1.0,1.0,1.0,1.0
g_BoxSize dd 1.5,1.0,1.5,1.0
g_BoxEdge dd 4 dup 0.03
g_ClearXYZ dd 0x00000000,0x00000000,0x00000000,0xffffffff
g_ClearW dd 0xffffffff,0xffffffff,0xffffffff,0x00000000
g_Quat dd 0.2,0.0,0.0,-1.0
IMG_END:
align 16
screen16 rb IMG_WIDTH*IMG_HEIGHT*16
align 16
screen4 rb IMG_WIDTH*IMG_HEIGHT*4
align 16
xxmm8:
rd 4
memStack:
rd 65536
align 16
I_END: