DOOM/linuxdoom-1.10

README - DOOM assembly code

Okay, I add the DOS assembly module for the historically
inclined here (may rec.games.programmer suffer). If anyone
feels the urge to port these to GNU GCC; either inline or
as separate modules including Makefile support, be my guest.

Module tmap.S includes the inner loops for texture mapping,
the interesting one being the floor/ceiling span rendering.

There was another module in the source dump, fpfunc.S, that
had both texture mapping and fixed point functions. It
contained implementations both for i386 and M68k. For
brevity, I include only the i386 fixed point stuff below.

//====================================================
// tmap.S as of January 10th, 1997

//================
//
// R_DrawColumn
//
//================

.data
loopcount .long 0
pixelcount .long 0

.text

.align 16
.globl _R_DrawColumn
_R_DrawColumn:

pushad

movl ebp,[_dc_yl]
movl ebx,ebp
movl edi,[_ylookup+ebx*4]
movl ebx,[_dc_x]
addl edi,[_columnofs + ebx*4]

movl eax,[_dc_yh]
incl eax
subl eax,ebp // pixel count
movl [pixelcount],eax // save for final pixel
js done // nothing to scale
shrl eax,1 // double pixel count
movl [loopcount],eax

movl ecx,[_dc_iscale]

movl eax,[_centery]
subl eax,ebp
imull ecx
movl ebp,[_dc_texturemid]
subl ebp,eax
shll ebp,9 // 7 significant bits, 25 frac

movl esi,[_dc_source]


movl ebx,[_dc_iscale]
shll ebx,9
movl eax,OFFSET patch1+2 // convice tasm to modify code...
movl [eax],ebx
movl eax,OFFSET patch2+2 // convice tasm to modify code...
movl [eax],ebx

// eax aligned colormap
// ebx aligned colormap
// ecx,edx scratch
// esi virtual source
// edi moving destination pointer
// ebp frac

movl ecx,ebp // begin calculating first pixel
addl ebp,ebx // advance frac pointer
shrl ecx,25 // finish calculation for first pixel
movl edx,ebp // begin calculating second pixel
addl ebp,ebx // advance frac pointer
shrl edx,25 // finish calculation for second pixel
movl eax,[_dc_colormap]
movl ebx,eax
movb al,[esi+ecx] // get first pixel
movb bl,[esi+edx] // get second pixel
movb al,[eax] // color translate first pixel
movb bl,[ebx] // color translate second pixel

testl [pixelcount],0fffffffeh
jnz doubleloop // at least two pixels to map
jmp checklast

.align 16
doubleloop:
movl ecx,ebp // begin calculating third pixel
patch1:
addl ebp,12345678h // advance frac pointer
movb [edi],al // write first pixel
shrl ecx,25 // finish calculation for third pixel
movl edx,ebp // begin calculating fourth pixel
patch2:
addl ebp,12345678h // advance frac pointer
movl [edi+SCREENWIDTH],bl // write second pixel
shrl edx,25 // finish calculation for fourth pixel
movb al,[esi+ecx] // get third pixel
addl edi,SCREENWIDTH*2 // advance to third pixel destination
movb bl,[esi+edx] // get fourth pixel
decl [loopcount] // done with loop?
movb al,[eax] // color translate third pixel
movb bl,[ebx] // color translate fourth pixel
jnz doubleloop

// check for final pixel
checklast:
testl [pixelcount],1
jz done
movb [edi],al // write final pixel

done:
popad
ret



//================
//
// R_DrawSpan
//
// Horizontal texture mapping
//
//================


.align 16
.globl _R_DrawSpan
_R_DrawSpan:
pushad

//
// find loop count
//
movl eax,[_ds_x2]
incl eax
subl eax,[_ds_x1] // pixel count
movl [pixelcount],eax // save for final pixel
js hdone // nothing to scale
shrl eax,1 // double pixel count
movl [loopcount],eax

//
// build composite position
//
movl ebp,[_ds_xfrac]
shll ebp,10
andl ebp,0ffff0000h
movl eax,[_ds_yfrac]
shrl eax,6
andl eax,0ffffh
orl ebp,eax

movl esi,[_ds_source]

//
// calculate screen dest
//
movl edi,[_ds_y]
movl edi,[_ylookup+edi*4]
movl eax,[_ds_x1]
addl edi,[_columnofs+eax*4]

//
// build composite step
//
movl ebx,[_ds_xstep]
shll ebx,10
andl ebx,0ffff0000h
movl eax,[_ds_ystep]
shrl eax,6
andl eax,0ffffh
orl ebx,eax

movl eax,OFFSET hpatch1+2 // convice tasm to modify code...
movl [eax],ebx
movl eax,OFFSET hpatch2+2 // convice tasm to modify code...
movl [eax],ebx

// eax aligned colormap
// ebx aligned colormap
// ecx,edx scratch
// esi virtual source
// edi moving destination pointer
// ebp frac

shldl ecx,ebp,22 // begin calculating third pixel (y units)
shldl ecx,ebp,6 // begin calculating third pixel (x units)
addl ebp,ebx // advance frac pointer
andl ecx,4095 // finish calculation for third pixel
shldl edx,ebp,22 // begin calculating fourth pixel (y units)
shldl edx,ebp,6 // begin calculating fourth pixel (x units)
addl ebp,ebx // advance frac pointer
andl edx,4095 // finish calculation for fourth pixel
movl eax,[_ds_colormap]
movl ebx,eax
movb al,[esi+ecx] // get first pixel
movb bl,[esi+edx] // get second pixel
movb al,[eax] // color translate first pixel
movb bl,[ebx] // color translate second pixel

testl [pixelcount],0fffffffeh
jnz hdoubleloop // at least two pixels to map
jmp hchecklast


.align 16
hdoubleloop:
shldl ecx,ebp,22 // begin calculating third pixel (y units)
shldl ecx,ebp,6 // begin calculating third pixel (x units)
hpatch1:
addl ebp,12345678h // advance frac pointer
movb [edi],al // write first pixel
andl ecx,4095 // finish calculation for third pixel
shldl edx,ebp,22 // begin calculating fourth pixel (y units)
shldl edx,ebp,6 // begin calculating fourth pixel (x units)
hpatch2:
addl ebp,12345678h // advance frac pointer
movb [edi+1],bl // write second pixel
andl edx,4095 // finish calculation for fourth pixel
movb al,[esi+ecx] // get third pixel
addl edi,2 // advance to third pixel destination
movb bl,[esi+edx] // get fourth pixel
decl [loopcount] // done with loop?
movb al,[eax] // color translate third pixel
movb bl,[ebx] // color translate fourth pixel
jnz hdoubleloop

// check for final pixel
hchecklast:
testl [pixelcount],1
jz hdone
movb [edi],al // write final pixel

hdone:
popad
ret




//====================================================
// fpfunc.S as of January 10th, 1997 (parts)

#ifdef i386

.text
.align 4
.globl _FixedMul
_FixedMul:
pushl %ebp
movl %esp,%ebp
movl 8(%ebp),%eax
imull 12(%ebp)
shrdl $16,%edx,%eax
popl %ebp
ret


.align 4
.globl _FixedDiv2
_FixedDiv2:
pushl %ebp
movl %esp,%ebp
movl 8(%ebp),%eax
cdq
shldl $16,%eax,%edx
sall $16,%eax
idivl 12(%ebp)
popl %ebp
ret

#endif