/*
THE COMPUTER CODE CONTAINED HEREIN IS THE SOLE PROPERTY OF PARALLAX
SOFTWARE CORPORATION ("PARALLAX").  PARALLAX, IN DISTRIBUTING THE CODE TO
END-USERS, AND SUBJECT TO ALL OF THE TERMS AND CONDITIONS HEREIN, GRANTS A
ROYALTY-FREE, PERPETUAL LICENSE TO SUCH END-USERS FOR USE BY SUCH END-USERS
IN USING, DISPLAYING,  AND CREATING DERIVATIVE WORKS THEREOF, SO LONG AS
SUCH USE, DISPLAY OR CREATION IS FOR NON-COMMERCIAL, ROYALTY OR REVENUE
FREE PURPOSES.  IN NO EVENT SHALL THE END-USER USE THE COMPUTER CODE
CONTAINED HEREIN FOR REVENUE-BEARING PURPOSES.  THE END-USER UNDERSTANDS
AND AGREES TO THE TERMS HEREIN AND ACCEPTS THE SAME BY USE OF THIS FILE.  
COPYRIGHT 1993-1999 PARALLAX SOFTWARE CORPORATION.  ALL RIGHTS RESERVED.
*/

#include "pa_enabl.h"                   //$$POLY_ACC
#include "mem.h"
#include "gr.h"
#include "grdef.h"
#include "rle.h"
#include "mono.h"
#include "byteswap.h"		// because of rle code that has short for row offsets
#include "error.h"

#if defined(POLY_ACC)
#include "poly_acc.h"
#endif

int gr_bitblt_dest_step_shift = 0;
int gr_bitblt_double = 0;
ubyte *gr_bitblt_fade_table=NULL;

extern void gr_vesa_bitmap( grs_bitmap * source, grs_bitmap * dest, int x, int y );

// This code aligns edi so that the destination is aligned to a dword boundry before rep movsd
void gr_linear_movsd(ubyte * src, ubyte * dest, int num_pixels );

#ifdef MACINTOSH

#define THRESHOLD	8

#ifdef RELEASE
#define test_byteblit	0
#else
ubyte test_byteblit = 0;
#endif

void gr_linear_movsd(ubyte * src, ubyte * dest, int num_pixels )
{
	int i;
	uint n, r;
	double *d, *s;
	ubyte *d1, *s1;

// check to see if we are starting on an even byte boundry
// if not, move appropriate number of bytes to even
// 8 byte boundry

	if ( (num_pixels < THRESHOLD) || (((int)src & 0x7) != ((int)dest & 0x7)) || test_byteblit ) {
		for (i = 0; i < num_pixels; i++)
			*dest++ = *src++;
		return;
	}

	i = 0;
	if ((r = (int)src & 0x7)) {
		for (i = 0; i < 8 - r; i++)
			*dest++ = *src++;
	}
	num_pixels -= i;

	n = num_pixels / 8;
	r = num_pixels % 8;
	s = (double *)src;
	d = (double *)dest;
	for (i = 0; i < n; i++)
		*d++ = *s++;
	s1 = (ubyte *)s;
	d1 = (ubyte *)d;
	for (i = 0; i < r; i++)
		*d1++ = *s1++;
}

#endif	//#ifdef MACINTOSH


void gr_linear_rep_movsdm(ubyte * src, ubyte * dest, int num_pixels );

#ifndef MACINTOSH
#pragma aux gr_linear_rep_movsdm parm [esi] [edi] [ecx] modify exact [ecx esi edi eax] = \
"nextpixel:"					\
	"mov	al,[esi]"			\
	"inc	esi"					\
	"cmp	al, 255"				\
	"je	skip_it"				\
	"mov	[edi], al"			\
"skip_it:"						\
	"inc	edi"					\
	"dec	ecx"					\
	"jne	nextpixel";

#else		// ifdef MACINTOSH

void gr_linear_rep_movsdm(ubyte * src, ubyte * dest, int num_pixels )
{
	int i;
	for (i=0; i<num_pixels; i++ )	{
		if (*src != TRANSPARENCY_COLOR )
			*dest = *src;
		dest++;
		src++;
	}
}

#endif		// ifdef MACINTOSH

void gr_linear_rep_movsdm_faded(ubyte * src, ubyte * dest, int num_pixels, ubyte fade_value );

#ifndef MACINTOSH
#pragma aux gr_linear_rep_movsdm_faded parm [esi] [edi] [ecx] [ebx] modify exact [ecx esi edi eax ebx] = \
"  xor eax, eax"	\
"  mov ah, bl"  \
"nextpixel:"					\
	"mov	al,[esi]"			\
	"inc	esi"					\
	"cmp	al, 255"				\
	"je	skip_it"				\
	"mov  al, gr_fade_table[eax]"	\
	"mov	[edi], al"			\
"skip_it:"						\
	"inc	edi"					\
	"dec	ecx"					\
	"jne	nextpixel";

#else

void gr_linear_rep_movsdm_faded(ubyte * src, ubyte * dest, int num_pixels, ubyte fade_value )
{
	int i;
	ubyte source;
	ubyte *fade_base;

	fade_base = gr_fade_table + (fade_value * 256);

	for (i=num_pixels; i != 0; i-- )
	{
		source = *src;
		if (source != (ubyte)TRANSPARENCY_COLOR )
			*dest = *(fade_base + source);
		dest++;
		src++;
	}
}

#endif		// ifdef MACINTOSH


void gr_linear_rep_movsd_2x(ubyte * src, ubyte * dest, int num_dest_pixels );

#ifndef MACINTOSH

#pragma aux gr_linear_rep_movsd_2x parm [esi] [edi] [ecx] modify exact [ecx esi edi eax ebx] = \
	"shr	ecx, 1"				\
	"jnc	nextpixel"			\
	"mov	al, [esi]"			\
	"mov	[edi], al"			\
	"inc	esi"					\
	"inc	edi"					\
	"cmp	ecx, 0"				\
	"je	done"					\
"nextpixel:"					\
	"mov	al,[esi]"			\
	"mov	ah, al"				\
	"mov	[edi], ax"			\
	"inc	esi"					\
	"inc	edi"					\
	"inc	edi"					\
	"dec	ecx"					\
	"jne	nextpixel"			\
"done:"

#else		// ifdef MACINTOSH

void gr_linear_rep_movsd_2x(ubyte *src, ubyte *dest, int num_pixels)
{
	double	*d = (double *)dest;
	uint	*s = (uint *)src;
	uint	doubletemp[2];
	uint	temp, work;
	int		i;

	if (num_pixels & 0x3) {					// not a multiple of 4?  do single pixel at a time
		for (i=0; i<num_pixels; i++) {
			*dest++ = *src;
			*dest++ = *src++;
		}
		return;
	}
	
	for (i = 0; i < num_pixels / 4; i++) {
		temp = work = *s++;

		temp = ((temp >> 8) & 0x00FFFF00) | (temp & 0xFF0000FF); // 0xABCDEFGH -> 0xABABCDEF
		temp = ((temp >> 8) & 0x000000FF) | (temp & 0xFFFFFF00); // 0xABABCDEF -> 0xABABCDCD
		doubletemp[0] = temp;

		work = ((work << 8) & 0x00FFFF00) | (work & 0xFF0000FF); // 0xABCDEFGH -> 0xABEFGHGH
		work = ((work << 8) & 0xFF000000) | (work & 0x00FFFFFF); // 0xABEFGHGH -> 0xEFEFGHGH
		doubletemp[1] = work;

		*d = *(double *) &(doubletemp[0]);
		d++;
	}
}

#endif

void modex_copy_column(ubyte * src, ubyte * dest, int num_pixels, int src_rowsize, int dest_rowsize );

#ifndef MACINTOSH
#pragma aux modex_copy_column parm [esi] [edi] [ecx] [ebx] [edx] modify exact [ecx esi edi] = \
"nextpixel:"							\
	"mov	al,[esi]"			\
	"add	esi, ebx"	\
	"mov	[edi], al"	\
	"add	edi, edx"	\
	"dec	ecx"			\
	"jne	nextpixel"
#else
void modex_copy_column(ubyte * src, ubyte * dest, int num_pixels, int src_rowsize, int dest_rowsize )
{
	src = src;
	dest = dest;
	num_pixels = num_pixels;
	src_rowsize = src_rowsize;
	dest_rowsize = dest_rowsize;
	Int3();
}
#endif

void modex_copy_column_m(ubyte * src, ubyte * dest, int num_pixels, int src_rowsize, int dest_rowsize );

#ifndef MACINTOSH

#pragma aux modex_copy_column_m parm [esi] [edi] [ecx] [ebx] [edx] modify exact [ecx esi edi] = \
"nextpixel:"							\
	"mov	al,[esi]"			\
	"add	esi, ebx"	\
	"cmp	al, 255"		\
	"je	skip_itx"		\
	"mov	[edi], al"	\
"skip_itx:"				\
	"add	edi, edx"	\
	"dec	ecx"			\
	"jne	nextpixel"
#else
void modex_copy_column_m(ubyte * src, ubyte * dest, int num_pixels, int src_rowsize, int dest_rowsize )
{
	src = src;
	dest = dest;
	num_pixels = num_pixels;
	src_rowsize = src_rowsize;
	dest_rowsize = dest_rowsize;
	Int3();
}
#endif

void gr_ubitmap00( int x, int y, grs_bitmap *bm )
{
	register int y1;
	int dest_rowsize;

	unsigned char * dest;
	unsigned char * src;

	dest_rowsize=grd_curcanv->cv_bitmap.bm_rowsize << gr_bitblt_dest_step_shift;
	dest = &(grd_curcanv->cv_bitmap.bm_data[ dest_rowsize*y+x ]);

	src = bm->bm_data;

	for (y1=0; y1 < bm->bm_h; y1++ )    {
		if (gr_bitblt_double)
			gr_linear_rep_movsd_2x( src, dest, bm->bm_w );
		else
			gr_linear_movsd( src, dest, bm->bm_w );
		src += bm->bm_rowsize;
		dest+= (int)(dest_rowsize);
	}
}

void gr_ubitmap00m( int x, int y, grs_bitmap *bm )
{
	register int y1;
	int dest_rowsize;

	unsigned char * dest;
	unsigned char * src;

	dest_rowsize=grd_curcanv->cv_bitmap.bm_rowsize << gr_bitblt_dest_step_shift;
	dest = &(grd_curcanv->cv_bitmap.bm_data[ dest_rowsize*y+x ]);

	src = bm->bm_data;

	if (gr_bitblt_fade_table==NULL)	{
		for (y1=0; y1 < bm->bm_h; y1++ )    {
			gr_linear_rep_movsdm( src, dest, bm->bm_w );
			src += bm->bm_rowsize;
			dest+= (int)(dest_rowsize);
		}
	} else {
		for (y1=0; y1 < bm->bm_h; y1++ )    {
			gr_linear_rep_movsdm_faded( src, dest, bm->bm_w, gr_bitblt_fade_table[y1+y] );
			src += bm->bm_rowsize;
			dest+= (int)(dest_rowsize);
		}
	}
}

//"		jmp 	aligned4				"	\
//"		mov	eax, edi				"	\
//"		and	eax, 11b				"	\
//"		jz		aligned4				"	\
//"		mov	ebx, 4				"	\
//"		sub	ebx, eax				"	\
//"		sub	ecx, ebx				"	\
//"alignstart:						"	\
//"		mov	al, [esi]			"	\
//"		add	esi, 4				"	\
//"		mov	[edi], al			"	\
//"		inc	edi					"	\
//"		dec	ebx					"	\
//"		jne	alignstart			"	\
//"aligned4:							"	\

void modex_copy_scanline( ubyte * src, ubyte * dest, int npixels );
#ifndef MACINTOSH
#pragma aux modex_copy_scanline parm [esi] [edi] [ecx] modify exact [ecx esi edi eax ebx edx] = \
"		mov	ebx, ecx				"	\
"		and	ebx, 11b				"	\
"		shr	ecx, 2				"	\
"		cmp	ecx, 0				"	\
"		je		no2group				"	\
"next4pixels:						"	\
"		mov	al, [esi+8]			"	\
"		mov	ah, [esi+12]		"	\
"		shl	eax, 16				"	\
"		mov	al, [esi]			"	\
"		mov	ah, [esi+4]			"	\
"		mov	[edi], eax			"	\
"		add	esi, 16				"	\
"		add	edi, 4				"	\
"		dec	ecx					"	\
"		jne	next4pixels			"	\
"no2group:							"	\
"		cmp	ebx, 0				"	\
"		je		done2					"	\
"finishend:							"	\
"		mov	al, [esi]			"	\
"		add	esi, 4				"	\
"		mov	[edi], al			"	\
"		inc	edi					"	\
"		dec	ebx					"	\
"		jne	finishend			"	\
"done2:								";
#else
void modex_copy_scanline( ubyte * src, ubyte * dest, int npixels )
{
	src = src;
	dest = dest;
	npixels = npixels;
	Int3();
}
#endif

void modex_copy_scanline_2x( ubyte * src, ubyte * dest, int npixels );

#ifndef MACINTOSH
#pragma aux modex_copy_scanline_2x parm [esi] [edi] [ecx] modify exact [ecx esi edi eax ebx edx] = \
"		mov	ebx, ecx				"	\
"		and	ebx, 11b				"	\
"		shr	ecx, 2				"	\
"		cmp	ecx, 0				"	\
"		je		no2group				"	\
"next4pixels:						"	\
"		mov	al, [esi+4]			"	\
"		mov	ah, [esi+6]			"	\
"		shl	eax, 16				"	\
"		mov	al, [esi]			"	\
"		mov	ah, [esi+2]			"	\
"		mov	[edi], eax			"	\
"		add	esi, 8				"	\
"		add	edi, 4				"	\
"		dec	ecx					"	\
"		jne	next4pixels			"	\
"no2group:							"	\
"		cmp	ebx, 0				"	\
"		je		done2					"	\
"finishend:							"	\
"		mov	al, [esi]			"	\
"		add	esi, 2				"	\
"		mov	[edi], al			"	\
"		inc	edi					"	\
"		dec	ebx					"	\
"		jne	finishend			"	\
"done2:								";
#else
void modex_copy_scanline_2x( ubyte * src, ubyte * dest, int npixels )
{
	src = src;
	dest = dest;
	npixels = npixels;
	Int3();
}
#endif


// From Linear to ModeX
void gr_bm_ubitblt01(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	ubyte * dbits;
	ubyte * sbits;
	int sstep,dstep;
	int y,plane;
	int w1;

	if ( w < 4 ) return;

	sstep = src->bm_rowsize;
	dstep = dest->bm_rowsize << gr_bitblt_dest_step_shift;

	if (!gr_bitblt_double)	{
		for (plane=0; plane<4; plane++ )	{
			gr_modex_setplane( (plane+dx)&3 );
			sbits = src->bm_data + (src->bm_rowsize * sy) + sx + plane;
			dbits = &gr_video_memory[(dest->bm_rowsize * dy) + ((plane+dx)/4) ];
			w1 = w >> 2;
			if ( (w&3) > plane ) w1++;
			for (y=dy; y < dy+h; y++ )		{
				modex_copy_scanline( sbits, dbits, w1 );
				dbits += dstep;
				sbits += sstep;
			}
		}
	} else {
		for (plane=0; plane<4; plane++ )	{
			gr_modex_setplane( (plane+dx)&3 );
			sbits = src->bm_data + (src->bm_rowsize * sy) + sx + plane/2;
			dbits = &gr_video_memory[(dest->bm_rowsize * dy) + ((plane+dx)/4) ];
			w1 = w >> 2;
			if ( (w&3) > plane ) w1++;
			for (y=dy; y < dy+h; y++ )		{
				modex_copy_scanline_2x( sbits, dbits, w1 );
				dbits += dstep;
				sbits += sstep;
			}
		}
	}
}


// From Linear to ModeX masked
void gr_bm_ubitblt01m(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	//ubyte * dbits1;
	//ubyte * sbits1;

	ubyte * dbits;
	ubyte * sbits;

	int x;
//	int y;

	sbits =   src->bm_data  + (src->bm_rowsize * sy) + sx;
	dbits =   &gr_video_memory[(dest->bm_rowsize * dy) + dx/4];

	for (x=dx; x < dx+w; x++ )	{
		gr_modex_setplane( x&3 );

		//sbits1 = sbits;
		//dbits1 = dbits;
		//for (y=0; y < h; y++ )    {
		//	*dbits1 = *sbits1;
		//	sbits1 += src_bm_rowsize;
		//	dbits1 += dest_bm_rowsize;
		//	}
		modex_copy_column_m(sbits, dbits, h, src->bm_rowsize, dest->bm_rowsize << gr_bitblt_dest_step_shift );

		sbits++;
		if ( (x&3)==3 )
			dbits++;
	}
}




void gr_ubitmap012( int x, int y, grs_bitmap *bm )
{
	register int x1, y1;
	unsigned char * src;

	src = bm->bm_data;

	for (y1=y; y1 < (y+bm->bm_h); y1++ )    {
		for (x1=x; x1 < (x+bm->bm_w); x1++ )    {
			gr_setcolor( *src++ );
			gr_upixel( x1, y1 );
		}
	}
}

void gr_ubitmap012m( int x, int y, grs_bitmap *bm )
{
	register int x1, y1;
	unsigned char * src;

	src = bm->bm_data;

	for (y1=y; y1 < (y+bm->bm_h); y1++ )    {
		for (x1=x; x1 < (x+bm->bm_w); x1++ )    {
			if ( *src != TRANSPARENCY_COLOR )	{
				gr_setcolor( *src );
				gr_upixel( x1, y1 );
			}
			src++;
		}
	}
}

#if defined(POLY_ACC)
void gr_ubitmap05( int x, int y, grs_bitmap *bm )
{
	register int x1, y1;
    unsigned char *src;
    short *dst;
    int mod;

    pa_flush();
    src = bm->bm_data;
    dst = (short *)(DATA + y * ROWSIZE + x * PA_BPP);
    mod = ROWSIZE / 2 - bm->bm_w;

	for (y1=y; y1 < (y+bm->bm_h); y1++ )    {
		for (x1=x; x1 < (x+bm->bm_w); x1++ )    {
            *dst++ = pa_clut[*src++];
        }
        dst += mod;
	}
}

void gr_ubitmap05m( int x, int y, grs_bitmap *bm )
{
	register int x1, y1;
    unsigned char *src;
    short *dst;
    int mod;

    pa_flush();
    src = bm->bm_data;
    dst = (short *)(DATA + y * ROWSIZE + x * PA_BPP);
    mod = ROWSIZE / 2 - bm->bm_w;

	for (y1=y; y1 < (y+bm->bm_h); y1++ )    {
		for (x1=x; x1 < (x+bm->bm_w); x1++ )    {
            if ( *src != TRANSPARENCY_COLOR )   {
                *dst = pa_clut[*src];
            }
			src++;
            ++dst;
		}
        dst += mod;
	}
}

void gr_bm_ubitblt05_rle(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
    unsigned short * dbits;
    unsigned char * sbits, scanline[640];
    int i, data_offset, j, nextrow;

    pa_flush();
    nextrow=dest->bm_rowsize/PA_BPP;

	data_offset = 1;
	if (src->bm_flags & BM_FLAG_RLE_BIG)
		data_offset = 2;

	sbits = &src->bm_data[4 + (src->bm_h*data_offset)];
	for (i=0; i<sy; i++ )
		sbits += (int)(INTEL_SHORT(src->bm_data[4+(i*data_offset)]));

    dbits = (unsigned short *)(dest->bm_data + (dest->bm_rowsize * dy) + dx*PA_BPP);

	// No interlacing, copy the whole buffer.
	for (i=0; i < h; i++ )    {
        gr_rle_expand_scanline( scanline, sbits, sx, sx+w-1 );
        for(j = 0; j != w; ++j)
            dbits[j] = pa_clut[scanline[j]];
        if ( src->bm_flags & BM_FLAG_RLE_BIG )
			sbits += (int)INTEL_SHORT(*((short *)&(src->bm_data[4+((i+sy)*data_offset)])));
	  	else
			sbits += (int)(src->bm_data[4+i+sy]);
        dbits += nextrow;
    }
}

void gr_bm_ubitblt05m_rle(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
    unsigned short * dbits;
    unsigned char * sbits, scanline[640];
    int i, data_offset, j, nextrow;

    pa_flush();
    nextrow=dest->bm_rowsize/PA_BPP;
	data_offset = 1;
	if (src->bm_flags & BM_FLAG_RLE_BIG)
		data_offset = 2;

	sbits = &src->bm_data[4 + (src->bm_h*data_offset)];
	for (i=0; i<sy; i++ )
		sbits += (int)(INTEL_SHORT(src->bm_data[4+(i*data_offset)]));

    dbits = (unsigned short *)(dest->bm_data + (dest->bm_rowsize * dy) + dx*PA_BPP);

	// No interlacing, copy the whole buffer.
	for (i=0; i < h; i++ )    {
        gr_rle_expand_scanline( scanline, sbits, sx, sx+w-1 );
        for(j = 0; j != w; ++j)
            if(scanline[j] != TRANSPARENCY_COLOR)
                dbits[j] = pa_clut[scanline[j]];
        if ( src->bm_flags & BM_FLAG_RLE_BIG )
			sbits += (int)INTEL_SHORT(*((short *)&(src->bm_data[4+((i+sy)*data_offset)])));
	  	else
			sbits += (int)(src->bm_data[4+i+sy]);
        dbits += nextrow;
	}
}
#endif

void gr_ubitmapGENERIC(int x, int y, grs_bitmap * bm)
{
	register int x1, y1;

	for (y1=0; y1 < bm->bm_h; y1++ )    {
		for (x1=0; x1 < bm->bm_w; x1++ )    {
			gr_setcolor( gr_gpixel(bm,x1,y1) );
			gr_upixel( x+x1, y+y1 );
		}
	}
}

void gr_ubitmapGENERICm(int x, int y, grs_bitmap * bm)
{
	register int x1, y1;
	ubyte c;

	for (y1=0; y1 < bm->bm_h; y1++ )    {
		for (x1=0; x1 < bm->bm_w; x1++ )    {
			c = gr_gpixel(bm,x1,y1);
			if ( c != TRANSPARENCY_COLOR )	{
				gr_setcolor( c );
				gr_upixel( x+x1, y+y1 );
			}
		}
	}
}



// From linear to SVGA
void gr_bm_ubitblt02(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	unsigned char * sbits;

	unsigned int offset, EndingOffset, VideoLocation;

	int sbpr, dbpr, y1, page, BytesToMove;

	sbpr = src->bm_rowsize;

	dbpr = dest->bm_rowsize << gr_bitblt_dest_step_shift;

	VideoLocation = (unsigned int)dest->bm_data + (dest->bm_rowsize * dy) + dx;

	sbits = src->bm_data + ( sbpr*sy ) + sx;

	for (y1=0; y1 < h; y1++ )    {

		page    = VideoLocation >> 16;
		offset  = VideoLocation & 0xFFFF;

		gr_vesa_setpage( page );

		EndingOffset = offset+w-1;

		if ( EndingOffset <= 0xFFFF )
		{
			if ( gr_bitblt_double )
				gr_linear_rep_movsd_2x( (void *)sbits, (void *)(offset+0xA0000), w );
			else
				gr_linear_movsd( (void *)sbits, (void *)(offset+0xA0000), w );

			VideoLocation += dbpr;
			sbits += sbpr;
		}
		else
		{
			BytesToMove = 0xFFFF-offset+1;

			if ( gr_bitblt_double )
				gr_linear_rep_movsd_2x( (void *)sbits, (void *)(offset+0xA0000), BytesToMove );
			else
				gr_linear_movsd( (void *)sbits, (void *)(offset+0xA0000), BytesToMove );

			page++;
			gr_vesa_setpage(page);

			if ( gr_bitblt_double )
				gr_linear_rep_movsd_2x( (void *)(sbits+BytesToMove/2), (void *)0xA0000, EndingOffset - 0xFFFF );
			else
				gr_linear_movsd( (void *)(sbits+BytesToMove), (void *)0xA0000, EndingOffset - 0xFFFF );

			VideoLocation += dbpr;
			sbits += sbpr;
		}
	}
}


void gr_bm_ubitblt02m(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	unsigned char * sbits;

	unsigned int offset, EndingOffset, VideoLocation;

	int sbpr, dbpr, y1, page, BytesToMove;

	sbpr = src->bm_rowsize;

	dbpr = dest->bm_rowsize << gr_bitblt_dest_step_shift;

	VideoLocation = (unsigned int)dest->bm_data + (dest->bm_rowsize * dy) + dx;

	sbits = src->bm_data + ( sbpr*sy ) + sx;

	for (y1=0; y1 < h; y1++ )    {

		page    = VideoLocation >> 16;
		offset  = VideoLocation & 0xFFFF;

		gr_vesa_setpage( page );

		EndingOffset = offset+w-1;

		if ( EndingOffset <= 0xFFFF )
		{
			gr_linear_rep_movsdm( (void *)sbits, (void *)(offset+0xA0000), w );

			VideoLocation += dbpr;
			sbits += sbpr;
		}
		else
		{
			BytesToMove = 0xFFFF-offset+1;

			gr_linear_rep_movsdm( (void *)sbits, (void *)(offset+0xA0000), BytesToMove );

			page++;
			gr_vesa_setpage(page);

			gr_linear_rep_movsdm( (void *)(sbits+BytesToMove), (void *)0xA0000, EndingOffset - 0xFFFF );

			VideoLocation += dbpr;
			sbits += sbpr;
		}
	}
}


// From SVGA to linear
void gr_bm_ubitblt20(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	unsigned char * dbits;

	unsigned int offset, offset1, offset2;

	int sbpr, dbpr, y1, page;

	dbpr = dest->bm_rowsize;

	sbpr = src->bm_rowsize;

	for (y1=0; y1 < h; y1++ )    {

		offset2 =   (unsigned int)src->bm_data  + (sbpr * (y1+sy)) + sx;
		dbits   =   dest->bm_data + (dbpr * (y1+dy)) + dx;

		page = offset2 >> 16;
		offset = offset2 & 0xFFFF;
		offset1 = offset+w-1;
		gr_vesa_setpage( page );

		if ( offset1 > 0xFFFF )  {
			// Overlaps two pages
			while( offset <= 0xFFFF )
				*dbits++ = gr_video_memory[offset++];
			offset1 -= (0xFFFF+1);
			offset = 0;
			page++;
			gr_vesa_setpage(page);
		}
		while( offset <= offset1 )
			*dbits++ = gr_video_memory[offset++];

	}
}

//@extern int Interlacing_on;

// From Linear to Linear
void gr_bm_ubitblt00(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	unsigned char * dbits;
	unsigned char * sbits;
	//int	src_bm_rowsize_2, dest_bm_rowsize_2;
	int dstep;

	int i;

	sbits =   src->bm_data  + (src->bm_rowsize * sy) + sx;
	dbits =   dest->bm_data + (dest->bm_rowsize * dy) + dx;

	dstep = dest->bm_rowsize << gr_bitblt_dest_step_shift;

	// No interlacing, copy the whole buffer.
	for (i=0; i < h; i++ )    {
		if (gr_bitblt_double)
			gr_linear_rep_movsd_2x( sbits, dbits, w );
		else
			gr_linear_movsd( sbits, dbits, w );
		sbits += src->bm_rowsize;
		dbits += dstep;
	}
}
// From Linear to Linear Masked
void gr_bm_ubitblt00m(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	unsigned char * dbits;
	unsigned char * sbits;
	//int	src_bm_rowsize_2, dest_bm_rowsize_2;

	int i;

	sbits =   src->bm_data  + (src->bm_rowsize * sy) + sx;
	dbits =   dest->bm_data + (dest->bm_rowsize * dy) + dx;

	// No interlacing, copy the whole buffer.

	if (gr_bitblt_fade_table==NULL)	{
		for (i=0; i < h; i++ )    {
			gr_linear_rep_movsdm( sbits, dbits, w );
			sbits += src->bm_rowsize;
			dbits += dest->bm_rowsize;
		}
	} else {
		for (i=0; i < h; i++ )    {
			gr_linear_rep_movsdm_faded( sbits, dbits, w, gr_bitblt_fade_table[dy+i] );
			sbits += src->bm_rowsize;
			dbits += dest->bm_rowsize;
		}
	}
}


extern void gr_lbitblt( grs_bitmap * source, grs_bitmap * dest, int height, int width );

#ifdef MACINTOSH

// width == number of destination pixels

void gr_linear_movsd_double(ubyte *src, ubyte *dest, int width)
{
	double	*d = (double *)dest;
	uint	*s = (uint *)src;
	uint	doubletemp[2];
	uint	temp, work;
	int		i, num_pixels;

	num_pixels = width / 2;
	if ( (num_pixels & 0x3) || (((int)src & 0x7) != ((int)dest & 0x7)) ) {					// not a multiple of 4?  do single pixel at a time
		for (i=0; i<num_pixels; i++) {
			*dest++ = *src;
			*dest++ = *src++;
		}
		return;
	}
	
	for (i = 0; i < num_pixels / 4; i++) {
		temp = work = *s++;

		temp = ((temp >> 8) & 0x00FFFF00) | (temp & 0xFF0000FF); // 0xABCDEFGH -> 0xABABCDEF
		temp = ((temp >> 8) & 0x000000FF) | (temp & 0xFFFFFF00); // 0xABABCDEF -> 0xABABCDCD
		doubletemp[0] = temp;

		work = ((work << 8) & 0x00FFFF00) | (work & 0xFF0000FF); // 0xABCDEFGH -> 0xABEFGHGH
		work = ((work << 8) & 0xFF000000) | (work & 0x00FFFFFF); // 0xABEFGHGH -> 0xEFEFGHGH
		doubletemp[1] = work;

		*d = *(double *) &(doubletemp[0]);
		d++;
	}
}

//extern void BlitLargeAlign(ubyte *draw_buffer, int dstRowBytes, ubyte *dstPtr, int w, int h, int modulus);

asm void BlitLargeAlign(ubyte *rSrcPtr, int rDblDStrd, ubyte *rDst1Ptr, int rWidth, int rHeight, int rModulus)
{
	stw		r31,-4(SP)				// store non-volatile reg in red zone
	addi	r5,r5,-8	// subtract 8 from dst
	stw		r30,-8(SP)				// store non-volatile reg in red zone

	la     	r30,-16(SP)		// calculate copy of local 8-byte variable
	sub     r9,r8,r6
											// rSStrd = modulus - w
	add		r31,r5,r4	// dst2 = dstRowBytes + dst1
	sub     r4,r4,r6			// r4 = dstRowBytes - w
	addi    r7,r7,-1	// subtract 1 from height count
	srawi   r6,r6,2		// rWidth = w >> 2
	addi	r3,r3,-4	// subtract 4 from src
	addi	r6,r6,-1		// subtract 1 from width count
	add     r4,r4,r4		// rDblDStrd = 2 * r4

BlitLargeAlignY:						// y count is in r7
	lwzu     r10,4(r3)			// load a long into r10
	mr       r0,r10  				// put a copy in r0
	mr       r11,r10
// these are simplified -- can't use 'em	inslwi   r0,r10,16,8
// these are simplified -- can't use 'em	insrwi   r11,r10,16,8
	rlwimi   r0,r10,24,8,31
	rlwimi   r11,r10,8,8,23
	rlwimi   r0,r10,16,24,31
	stw      r0,0(r30)
	rlwimi   r11,r10,16,0,7
	stw      r11,4(r30)
	mtctr	 	r6				// copy x count into the counter
	lfd      fp0,0(r30)

BlitLargeAlignX:
	lwzu     r10,4(r3)			// load a long into r10
	stfdu    fp0,8(r5)
	mr       r0,r10  				// put a copy in r0
	mr       r11,r10
// simplefied	inslwi   r0,r10,16,8
// simplefied	insrwi   r11,r10,16,8
	rlwimi   r0,r10,24,8,31
	rlwimi   r11,r10,8,8,23
	rlwimi   r0,r10,16,24,31
	stw      r0,0(r30)
	rlwimi   r11,r10,16,0,7
	stw      r11,4(r30)
	stfdu    fp0,8(r31)
	lfd      fp0,0(r30)
	bdnz	 BlitLargeAlignX		// loop over all x

	stfdu    fp0,8(r5)
	addic.   r7,r7,-1		// decrement the counter
	add      r3,r3,r9
											// src += sstride
	add      r5,r5,r4
											// dst1 += dstride
	stfdu    fp0,8(r31)
	add      r31,r31,r4
											// dst2 += dstride
	bne      BlitLargeAlignY      // loop for all y

	lwz		r30,-8(SP)				// restore non-volatile regs
	lwz		r31,-4(SP)				// restore non-volatile regs
	blr								// return to caller
}

void gr_bm_ubitblt_double(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap *src, grs_bitmap *dest)
{
	ubyte * dbits;
	ubyte * sbits;
	int dstep, i;

	sbits = src->bm_data  + (src->bm_rowsize * sy) + sx;
	dbits = dest->bm_data + (dest->bm_rowsize * dy) + dx;
	dstep = dest->bm_rowsize << gr_bitblt_dest_step_shift;
	Assert( !((int)dbits & 0x7) );		// assert to check double word alignment
	BlitLargeAlign(sbits, dstep, dbits, src->bm_w, src->bm_h, src->bm_rowsize);
}

// w and h are the doubled width and height

void gr_bm_ubitblt_double_slow(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap *src, grs_bitmap *dest)
{
	ubyte * dbits;
	ubyte * sbits;
	int dstep, i, j;

	sbits = src->bm_data  + (src->bm_rowsize * sy) + sx;
	dbits = dest->bm_data + (dest->bm_rowsize * dy) + dx;
	dstep = dest->bm_rowsize << gr_bitblt_dest_step_shift;

	for (i=0; i < h; i++ )    {

		gr_linear_movsd_double(sbits, dbits, w);
		dbits += dstep;
		if (i & 1)
			sbits += src->bm_rowsize;
	}
}

#endif


// Clipped bitmap ...

void gr_bitmap( int x, int y, grs_bitmap *bm )
{
	int dx1=x, dx2=x+bm->bm_w-1;
	int dy1=y, dy2=y+bm->bm_h-1;
	int sx=0, sy=0;

	if ((dx1 >= grd_curcanv->cv_bitmap.bm_w ) || (dx2 < 0)) return;
	if ((dy1 >= grd_curcanv->cv_bitmap.bm_h) || (dy2 < 0)) return;
	if ( dx1 < 0 ) { sx = -dx1; dx1 = 0; }
	if ( dy1 < 0 ) { sy = -dy1; dy1 = 0; }
	if ( dx2 >= grd_curcanv->cv_bitmap.bm_w )	{ dx2 = grd_curcanv->cv_bitmap.bm_w-1; }
	if ( dy2 >= grd_curcanv->cv_bitmap.bm_h )	{ dy2 = grd_curcanv->cv_bitmap.bm_h-1; }

	// Draw bitmap bm[x,y] into (dx1,dy1)-(dx2,dy2)

	gr_bm_ubitblt(dx2-dx1+1,dy2-dy1+1, dx1, dy1, sx, sy, bm, &grd_curcanv->cv_bitmap );

}

//-NOT-used // From linear to SVGA
//-NOT-used void gr_bm_ubitblt02_2x(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
//-NOT-used {
//-NOT-used 	unsigned char * sbits;
//-NOT-used
//-NOT-used 	unsigned int offset, EndingOffset, VideoLocation;
//-NOT-used
//-NOT-used 	int sbpr, dbpr, y1, page, BytesToMove;
//-NOT-used
//-NOT-used 	sbpr = src->bm_rowsize;
//-NOT-used
//-NOT-used 	dbpr = dest->bm_rowsize << gr_bitblt_dest_step_shift;
//-NOT-used
//-NOT-used 	VideoLocation = (unsigned int)dest->bm_data + (dest->bm_rowsize * dy) + dx;
//-NOT-used
//-NOT-used 	sbits = src->bm_data + ( sbpr*sy ) + sx;
//-NOT-used
//-NOT-used 	for (y1=0; y1 < h; y1++ )    {
//-NOT-used
//-NOT-used 		page    = VideoLocation >> 16;
//-NOT-used 		offset  = VideoLocation & 0xFFFF;
//-NOT-used
//-NOT-used 		gr_vesa_setpage( page );
//-NOT-used
//-NOT-used 		EndingOffset = offset+w-1;
//-NOT-used
//-NOT-used 		if ( EndingOffset <= 0xFFFF )
//-NOT-used 		{
//-NOT-used 			gr_linear_rep_movsd_2x( (void *)sbits, (void *)(offset+0xA0000), w );
//-NOT-used
//-NOT-used 			VideoLocation += dbpr;
//-NOT-used 			sbits += sbpr;
//-NOT-used 		}
//-NOT-used 		else
//-NOT-used 		{
//-NOT-used 			BytesToMove = 0xFFFF-offset+1;
//-NOT-used
//-NOT-used 			gr_linear_rep_movsd_2x( (void *)sbits, (void *)(offset+0xA0000), BytesToMove );
//-NOT-used
//-NOT-used 			page++;
//-NOT-used 			gr_vesa_setpage(page);
//-NOT-used
//-NOT-used 			gr_linear_rep_movsd_2x( (void *)(sbits+BytesToMove/2), (void *)0xA0000, EndingOffset - 0xFFFF );
//-NOT-used
//-NOT-used 			VideoLocation += dbpr;
//-NOT-used 			sbits += sbpr;
//-NOT-used 		}
//-NOT-used
//-NOT-used
//-NOT-used 	}
//-NOT-used }


//-NOT-used // From Linear to Linear
//-NOT-used void gr_bm_ubitblt00_2x(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
//-NOT-used {
//-NOT-used 	unsigned char * dbits;
//-NOT-used 	unsigned char * sbits;
//-NOT-used 	//int	src_bm_rowsize_2, dest_bm_rowsize_2;
//-NOT-used
//-NOT-used 	int i;
//-NOT-used
//-NOT-used 	sbits =   src->bm_data  + (src->bm_rowsize * sy) + sx;
//-NOT-used 	dbits =   dest->bm_data + (dest->bm_rowsize * dy) + dx;
//-NOT-used
//-NOT-used 	// No interlacing, copy the whole buffer.
//-NOT-used 	for (i=0; i < h; i++ )    {
//-NOT-used 		gr_linear_rep_movsd_2x( sbits, dbits, w );
//-NOT-used
//-NOT-used 		sbits += src->bm_rowsize;
//-NOT-used 		dbits += dest->bm_rowsize << gr_bitblt_dest_step_shift;
//-NOT-used 	}
//-NOT-used }

void gr_bm_ubitblt00_rle(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	unsigned char * dbits;
	unsigned char * sbits;
	int i, data_offset;

	data_offset = 1;
	if (src->bm_flags & BM_FLAG_RLE_BIG)
		data_offset = 2;

	sbits = &src->bm_data[4 + (src->bm_h*data_offset)];

	for (i=0; i<sy; i++ )
		sbits += (int)(INTEL_SHORT(src->bm_data[4+(i*data_offset)]));

	dbits = dest->bm_data + (dest->bm_rowsize * dy) + dx;

	// No interlacing, copy the whole buffer.
	for (i=0; i < h; i++ )    {
		gr_rle_expand_scanline( dbits, sbits, sx, sx+w-1 );
		if ( src->bm_flags & BM_FLAG_RLE_BIG )
			sbits += (int)INTEL_SHORT(*((short *)&(src->bm_data[4+((i+sy)*data_offset)])));
	  	else
			sbits += (int)(src->bm_data[4+i+sy]);
		dbits += dest->bm_rowsize << gr_bitblt_dest_step_shift;
	}
}

void gr_bm_ubitblt00m_rle(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	unsigned char * dbits;
	unsigned char * sbits;
	int i, data_offset;

	data_offset = 1;
	if (src->bm_flags & BM_FLAG_RLE_BIG)
		data_offset = 2;

	sbits = &src->bm_data[4 + (src->bm_h*data_offset)];
	for (i=0; i<sy; i++ )
		sbits += (int)(INTEL_SHORT(src->bm_data[4+(i*data_offset)]));

	dbits = dest->bm_data + (dest->bm_rowsize * dy) + dx;

	// No interlacing, copy the whole buffer.
	for (i=0; i < h; i++ )    {
		gr_rle_expand_scanline_masked( dbits, sbits, sx, sx+w-1 );
		if ( src->bm_flags & BM_FLAG_RLE_BIG )
			sbits += (int)INTEL_SHORT(*((short *)&(src->bm_data[4+((i+sy)*data_offset)])));
	  	else
			sbits += (int)(src->bm_data[4+i+sy]);
		dbits += dest->bm_rowsize << gr_bitblt_dest_step_shift;
	}
}

// in rle.c

extern void gr_rle_expand_scanline_generic( grs_bitmap * dest, int dx, int dy, ubyte *src, int x1, int x2  );
extern void gr_rle_expand_scanline_generic_masked( grs_bitmap * dest, int dx, int dy, ubyte *src, int x1, int x2  );


void gr_bm_ubitblt0x_rle(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	int i, data_offset;
	register int y1;
	unsigned char * sbits;

	//mprintf( 0, "SVGA RLE!\n" );

	data_offset = 1;
	if (src->bm_flags & BM_FLAG_RLE_BIG)
		data_offset = 2;

	sbits = &src->bm_data[4 + (src->bm_h*data_offset)];
	for (i=0; i<sy; i++ )
		sbits += (int)(INTEL_SHORT(src->bm_data[4+(i*data_offset)]));

	for (y1=0; y1 < h; y1++ )    {
		gr_rle_expand_scanline_generic( dest, dx, dy+y1,  sbits, sx, sx+w-1  );
		if ( src->bm_flags & BM_FLAG_RLE_BIG )
			sbits += (int)INTEL_SHORT(*((short *)&(src->bm_data[4+((y1+sy)*data_offset)])));
		else
			sbits += (int)src->bm_data[4+y1+sy];
	}

}

void gr_bm_ubitblt0xm_rle(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	int i, data_offset;
	register int y1;
	unsigned char * sbits;

	//mprintf( 0, "SVGA RLE!\n" );

	data_offset = 1;
	if (src->bm_flags & BM_FLAG_RLE_BIG)
		data_offset = 2;

	sbits = &src->bm_data[4 + (src->bm_h*data_offset)];
	for (i=0; i<sy; i++ )
		sbits += (int)(INTEL_SHORT(src->bm_data[4+(i*data_offset)]));

	for (y1=0; y1 < h; y1++ )    {
		gr_rle_expand_scanline_generic_masked( dest, dx, dy+y1,  sbits, sx, sx+w-1  );
		if ( src->bm_flags & BM_FLAG_RLE_BIG )
			sbits += (int)INTEL_SHORT(*((short *)&(src->bm_data[4+((y1+sy)*data_offset)])));
		else
			sbits += (int)src->bm_data[4+y1+sy];
	}

}


void gr_bm_ubitblt02m_rle(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	int i, data_offset;
	register int y1;
	unsigned char * sbits;

	//mprintf( 0, "SVGA RLE!\n" );

	data_offset = 1;
	if (src->bm_flags & BM_FLAG_RLE_BIG)
		data_offset = 2;

	sbits = &src->bm_data[4 + (src->bm_h*data_offset)];
	for (i=0; i<sy; i++ )
		sbits += (int)(INTEL_SHORT(src->bm_data[4+(i*data_offset)]));

	for (y1=0; y1 < h; y1++ )    {
		gr_rle_expand_scanline_svga_masked( dest, dx, dy+y1,  sbits, sx, sx+w-1  );
		if ( src->bm_flags & BM_FLAG_RLE_BIG )
			sbits += (int)INTEL_SHORT(*((short *)&(src->bm_data[4+((y1+sy)*data_offset)])));
		else
			sbits += (int)src->bm_data[4+y1+sy];
	}
}


void gr_bm_ubitblt(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	register int x1, y1;

	if ( (src->bm_type == BM_LINEAR) && (dest->bm_type == BM_LINEAR ))
	{
		if ( src->bm_flags & BM_FLAG_RLE )
			gr_bm_ubitblt00_rle( w, h, dx, dy, sx, sy, src, dest );
		else
			gr_bm_ubitblt00( w, h, dx, dy, sx, sy, src, dest );
		return;
	}

	if ( (src->bm_flags & BM_FLAG_RLE ) && (src->bm_type == BM_LINEAR) )	{
		gr_bm_ubitblt0x_rle(w, h, dx, dy, sx, sy, src, dest );
	 	return;
	}

	if ( (src->bm_type == BM_LINEAR) && (dest->bm_type == BM_SVGA ))
	{
		gr_bm_ubitblt02( w, h, dx, dy, sx, sy, src, dest );
		return;
	}

	if ( (src->bm_type == BM_SVGA) && (dest->bm_type == BM_LINEAR ))
	{
		gr_bm_ubitblt20( w, h, dx, dy, sx, sy, src, dest );
		return;
	}

	if ( (src->bm_type == BM_LINEAR) && (dest->bm_type == BM_MODEX ))
	{
		gr_bm_ubitblt01( w, h, dx+XOFFSET, dy+YOFFSET, sx, sy, src, dest );
		return;
	}

#if defined(POLY_ACC)
    if ( (src->bm_type == BM_LINEAR) && (dest->bm_type == BM_LINEAR15 ))
	{
        ubyte *s = src->bm_data + sy * src->bm_rowsize + sx;
        ushort *t = (ushort *)(dest->bm_data + dy * dest->bm_rowsize + dx * PA_BPP);
        int x;
        pa_flush();
        for(;h--;)
        {
            for(x = 0; x < w; x++)
                t[x] = pa_clut[s[x]];
            s += src->bm_rowsize;
            t += dest->bm_rowsize / PA_BPP;
        }
        return;
	}

    if ( (src->bm_type == BM_LINEAR15) && (dest->bm_type == BM_LINEAR15 ))
	{
        pa_blit(dest, dx, dy, src, sx, sy, w, h);
        return;
	}
#endif

    for (y1=0; y1 < h; y1++ )    {
		for (x1=0; x1 < w; x1++ )    {
			gr_bm_pixel( dest, dx+x1, dy+y1, gr_gpixel(src,sx+x1,sy+y1) );
		}
	}
}

void gr_bm_bitblt(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	int dx1=dx, dx2=dx+dest->bm_w-1;
	int dy1=dy, dy2=dy+dest->bm_h-1;

	int sx1=sx, sx2=sx+src->bm_w-1;
	int sy1=sy, sy2=sy+src->bm_h-1;

	if ((dx1 >= dest->bm_w ) || (dx2 < 0)) return;
	if ((dy1 >= dest->bm_h ) || (dy2 < 0)) return;
	if ( dx1 < 0 ) { sx1 += -dx1; dx1 = 0; }
	if ( dy1 < 0 ) { sy1 += -dy1; dy1 = 0; }
	if ( dx2 >= dest->bm_w )	{ dx2 = dest->bm_w-1; }
	if ( dy2 >= dest->bm_h )	{ dy2 = dest->bm_h-1; }

	if ((sx1 >= src->bm_w ) || (sx2 < 0)) return;
	if ((sy1 >= src->bm_h ) || (sy2 < 0)) return;
	if ( sx1 < 0 ) { dx1 += -sx1; sx1 = 0; }
	if ( sy1 < 0 ) { dy1 += -sy1; sy1 = 0; }
	if ( sx2 >= src->bm_w )	{ sx2 = src->bm_w-1; }
	if ( sy2 >= src->bm_h )	{ sy2 = src->bm_h-1; }

	// Draw bitmap bm[x,y] into (dx1,dy1)-(dx2,dy2)
	if ( dx2-dx1+1 < w )
		w = dx2-dx1+1;
	if ( dy2-dy1+1 < h )
		h = dy2-dy1+1;
	if ( sx2-sx1+1 < w )
		w = sx2-sx1+1;
	if ( sy2-sy1+1 < h )
		h = sy2-sy1+1;

	gr_bm_ubitblt(w,h, dx1, dy1, sx1, sy1, src, dest );
}

void gr_ubitmap( int x, int y, grs_bitmap *bm )
{   int source, dest;

	source = bm->bm_type;
	dest = TYPE;

	if (source==BM_LINEAR) {
		switch( dest )
		{
		case BM_LINEAR:
			if ( bm->bm_flags & BM_FLAG_RLE )
				gr_bm_ubitblt00_rle(bm->bm_w, bm->bm_h, x, y, 0, 0, bm, &grd_curcanv->cv_bitmap );
			else
				gr_ubitmap00( x, y, bm );
			return;
		case BM_SVGA:
			if ( bm->bm_flags & BM_FLAG_RLE )
				gr_bm_ubitblt0x_rle(bm->bm_w, bm->bm_h, x, y, 0, 0, bm, &grd_curcanv->cv_bitmap );
			else
				gr_vesa_bitmap( bm, &grd_curcanv->cv_bitmap, x, y );
			return;
		case BM_MODEX:
			gr_bm_ubitblt01(bm->bm_w, bm->bm_h, x+XOFFSET, y+YOFFSET, 0, 0, bm, &grd_curcanv->cv_bitmap);
			return;
#if defined(POLY_ACC)
        case BM_LINEAR15:
            if ( bm->bm_flags & BM_FLAG_RLE )
                gr_bm_ubitblt05_rle(bm->bm_w, bm->bm_h, x, y, 0, 0, bm, &grd_curcanv->cv_bitmap );
            else
                gr_ubitmap05( x, y, bm);
			return;

#endif
		default:
			gr_ubitmap012( x, y, bm );
			return;
		}
	} else  {
		gr_ubitmapGENERIC(x, y, bm);
	}
}


void gr_ubitmapm( int x, int y, grs_bitmap *bm )
{   int source, dest;

	source = bm->bm_type;
	dest = TYPE;

	Assert(x+bm->bm_w <= grd_curcanv->cv_w);
	Assert(y+bm->bm_h <= grd_curcanv->cv_h);

   #ifdef _3DFX
   _3dfx_Blit( x, y, bm );
   if ( _3dfx_skip_ddraw )
      return;
   #endif

	if (source==BM_LINEAR) {
		switch( dest )
		{
		case BM_LINEAR:
			if ( bm->bm_flags & BM_FLAG_RLE )
				gr_bm_ubitblt00m_rle(bm->bm_w, bm->bm_h, x, y, 0, 0, bm, &grd_curcanv->cv_bitmap );
			else
				gr_ubitmap00m( x, y, bm );
			return;
		case BM_SVGA:
			if (bm->bm_flags & BM_FLAG_RLE)
				gr_bm_ubitblt02m_rle(bm->bm_w, bm->bm_h, x, y, 0, 0, bm, &grd_curcanv->cv_bitmap);
				//gr_bm_ubitblt0xm_rle(bm->bm_w, bm->bm_h, x, y, 0, 0, bm, &grd_curcanv->cv_bitmap);
			else
				gr_bm_ubitblt02m(bm->bm_w, bm->bm_h, x, y, 0, 0, bm, &grd_curcanv->cv_bitmap);
			//gr_ubitmapGENERICm(x, y, bm);
			return;
		case BM_MODEX:
			gr_bm_ubitblt01m(bm->bm_w, bm->bm_h, x+XOFFSET, y+YOFFSET, 0, 0, bm, &grd_curcanv->cv_bitmap);
			return;
#if defined(POLY_ACC)
        case BM_LINEAR15:
            if ( bm->bm_flags & BM_FLAG_RLE )
                gr_bm_ubitblt05m_rle(bm->bm_w, bm->bm_h, x, y, 0, 0, bm, &grd_curcanv->cv_bitmap );
            else
                gr_ubitmap05m( x, y, bm );
            return;
#endif

		default:
			gr_ubitmap012m( x, y, bm );
			return;
		}
	} else  {
		gr_ubitmapGENERICm(x, y, bm);
	}
}


void gr_bitmapm( int x, int y, grs_bitmap *bm )
{
	int dx1=x, dx2=x+bm->bm_w-1;
	int dy1=y, dy2=y+bm->bm_h-1;
	int sx=0, sy=0;

	if ((dx1 >= grd_curcanv->cv_bitmap.bm_w ) || (dx2 < 0)) return;
	if ((dy1 >= grd_curcanv->cv_bitmap.bm_h) || (dy2 < 0)) return;
	if ( dx1 < 0 ) { sx = -dx1; dx1 = 0; }
	if ( dy1 < 0 ) { sy = -dy1; dy1 = 0; }
	if ( dx2 >= grd_curcanv->cv_bitmap.bm_w )	{ dx2 = grd_curcanv->cv_bitmap.bm_w-1; }
	if ( dy2 >= grd_curcanv->cv_bitmap.bm_h )	{ dy2 = grd_curcanv->cv_bitmap.bm_h-1; }

	// Draw bitmap bm[x,y] into (dx1,dy1)-(dx2,dy2)

	if ( (bm->bm_type == BM_LINEAR) && (grd_curcanv->cv_bitmap.bm_type == BM_LINEAR ))
	{
		if ( bm->bm_flags & BM_FLAG_RLE )
			gr_bm_ubitblt00m_rle(dx2-dx1+1,dy2-dy1+1, dx1, dy1, sx, sy, bm, &grd_curcanv->cv_bitmap );
		else
			gr_bm_ubitblt00m(dx2-dx1+1,dy2-dy1+1, dx1, dy1, sx, sy, bm, &grd_curcanv->cv_bitmap );
		return;
	}
	else if ( (bm->bm_type == BM_LINEAR) && (grd_curcanv->cv_bitmap.bm_type == BM_SVGA ))
	{
		gr_bm_ubitblt02m(dx2-dx1+1,dy2-dy1+1, dx1, dy1, sx, sy, bm, &grd_curcanv->cv_bitmap );
		return;
	}

	gr_bm_ubitbltm(dx2-dx1+1,dy2-dy1+1, dx1, dy1, sx, sy, bm, &grd_curcanv->cv_bitmap );

}

void gr_bm_ubitbltm(int w, int h, int dx, int dy, int sx, int sy, grs_bitmap * src, grs_bitmap * dest)
{
	register int x1, y1;
    ubyte c;

#if defined(POLY_ACC)
    if(src->bm_type == BM_LINEAR && dest->bm_type == BM_LINEAR15)
    {
        ubyte *s;
        ushort *d;
        ushort u;
        int smod, dmod;

        pa_flush();
        s = (ubyte *)(src->bm_data + src->bm_rowsize * sy + sx);
        smod = src->bm_rowsize - w;
        d = (ushort *)(dest->bm_data + dest->bm_rowsize * dy + dx * PA_BPP);
        dmod = dest->bm_rowsize / PA_BPP - w;
        for (; h--;) {
            for (x1=w; x1--; ) {
                if ((u = *s) != TRANSPARENCY_COLOR)
                    *d = pa_clut[u];
                ++s;
                ++d;
            }
            s += smod;
            d += dmod;
        }
    }

    if(src->bm_type == BM_LINEAR15)
    {
        Assert(src->bm_type == dest->bm_type);         // I don't support 15 to 8 yet.
        pa_blit_transparent(dest, dx, dy, src, sx, sy, w, h);
        return;
    }
#endif

    for (y1=0; y1 < h; y1++ )    {
		for (x1=0; x1 < w; x1++ )    {
			if ((c=gr_gpixel(src,sx+x1,sy+y1))!=TRANSPARENCY_COLOR)
				gr_bm_pixel( dest, dx+x1, dy+y1,c  );
		}
	}
}