r300_emit.c

Mesa is an open-source implementation of the OpenGL specification - a system for rendering interacti

/*
Copyright (C) The Weather Channel, Inc.  2002.  All Rights Reserved.

The Weather Channel (TM) funded Tungsten Graphics to develop the
initial release of the Radeon 8500 driver under the XFree86 license.
This notice must be preserved.

Permission is hereby granted, free of charge, to any person obtaining
a copy of this software and associated documentation files (the
"Software"), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish,
distribute, sublicense, and/or sell copies of the Software, and to
permit persons to whom the Software is furnished to do so, subject to
the following conditions:

The above copyright notice and this permission notice (including the
next paragraph) shall be included in all copies or substantial
portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
IN NO EVENT SHALL THE COPYRIGHT OWNER(S) AND/OR ITS SUPPLIERS BE
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

**************************************************************************/

/**
 * \file
 *
 * \author Keith Whitwell <[e-mail]>
 */

#include "glheader.h"
#include "mtypes.h"
#include "colormac.h"
#include "imports.h"
#include "macros.h"
#include "image.h"

#include "swrast_setup/swrast_setup.h"
#include "math/m_translate.h"
#include "tnl/tnl.h"
#include "tnl/t_context.h"

#include "r300_context.h"
#include "radeon_ioctl.h"
#include "r300_state.h"
#include "r300_emit.h"
#include "r300_ioctl.h"

#ifdef USER_BUFFERS
#include "r300_mem.h"
#endif

#if SWIZZLE_X != R300_INPUT_ROUTE_SELECT_X || \
    SWIZZLE_Y != R300_INPUT_ROUTE_SELECT_Y || \
    SWIZZLE_Z != R300_INPUT_ROUTE_SELECT_Z || \
    SWIZZLE_W != R300_INPUT_ROUTE_SELECT_W || \
    SWIZZLE_ZERO != R300_INPUT_ROUTE_SELECT_ZERO || \
    SWIZZLE_ONE != R300_INPUT_ROUTE_SELECT_ONE
#error Cannot change these!
#endif

#define DEBUG_ALL DEBUG_VERTS

#if defined(USE_X86_ASM)
#define COPY_DWORDS( dst, src, nr )					\
do {									\
	int __tmp;							\
	__asm__ __volatile__( "rep ; movsl"				\
			      : "=%c" (__tmp), "=D" (dst), "=S" (__tmp)	\
			      : "0" (nr),				\
			        "D" ((long)dst),			\
			        "S" ((long)src) );			\
} while (0)
#else
#define COPY_DWORDS( dst, src, nr )		\
do {						\
   int j;					\
   for ( j = 0 ; j < nr ; j++ )			\
      dst[j] = ((int *)src)[j];			\
   dst += nr;					\
} while (0)
#endif

static void r300EmitVec4(GLcontext * ctx, struct r300_dma_region *rvb,
			 GLvoid * data, int stride, int count)
{
	int i;
	int *out = (int *)(rvb->address + rvb->start);

	if (RADEON_DEBUG & DEBUG_VERTS)
		fprintf(stderr, "%s count %d stride %d out %p data %p\n",
			__FUNCTION__, count, stride, (void *)out, (void *)data);

	if (stride == 4)
		COPY_DWORDS(out, data, count);
	else
		for (i = 0; i < count; i++) {
			out[0] = *(int *)data;
			out++;
			data += stride;
		}
}

static void r300EmitVec8(GLcontext * ctx, struct r300_dma_region *rvb,
			 GLvoid * data, int stride, int count)
{
	int i;
	int *out = (int *)(rvb->address + rvb->start);

	if (RADEON_DEBUG & DEBUG_VERTS)
		fprintf(stderr, "%s count %d stride %d out %p data %p\n",
			__FUNCTION__, count, stride, (void *)out, (void *)data);

	if (stride == 8)
		COPY_DWORDS(out, data, count * 2);
	else
		for (i = 0; i < count; i++) {
			out[0] = *(int *)data;
			out[1] = *(int *)(data + 4);
			out += 2;
			data += stride;
		}
}

static void r300EmitVec12(GLcontext * ctx, struct r300_dma_region *rvb,
			  GLvoid * data, int stride, int count)
{
	int i;
	int *out = (int *)(rvb->address + rvb->start);

	if (RADEON_DEBUG & DEBUG_VERTS)
		fprintf(stderr, "%s count %d stride %d out %p data %p\n",
			__FUNCTION__, count, stride, (void *)out, (void *)data);

	if (stride == 12)
		COPY_DWORDS(out, data, count * 3);
	else
		for (i = 0; i < count; i++) {
			out[0] = *(int *)data;
			out[1] = *(int *)(data + 4);
			out[2] = *(int *)(data + 8);
			out += 3;
			data += stride;
		}
}

static void r300EmitVec16(GLcontext * ctx, struct r300_dma_region *rvb,
			  GLvoid * data, int stride, int count)
{
	int i;
	int *out = (int *)(rvb->address + rvb->start);

	if (RADEON_DEBUG & DEBUG_VERTS)
		fprintf(stderr, "%s count %d stride %d out %p data %p\n",
			__FUNCTION__, count, stride, (void *)out, (void *)data);

	if (stride == 16)
		COPY_DWORDS(out, data, count * 4);
	else
		for (i = 0; i < count; i++) {
			out[0] = *(int *)data;
			out[1] = *(int *)(data + 4);
			out[2] = *(int *)(data + 8);
			out[3] = *(int *)(data + 12);
			out += 4;
			data += stride;
		}
}

static void r300EmitVec(GLcontext * ctx, struct r300_dma_region *rvb,
			GLvoid * data, int size, int stride, int count)
{
	r300ContextPtr rmesa = R300_CONTEXT(ctx);

	if (stride == 0) {
		r300AllocDmaRegion(rmesa, rvb, size * 4, 4);
		count = 1;
		rvb->aos_offset = GET_START(rvb);
		rvb->aos_stride = 0;
	} else {
		r300AllocDmaRegion(rmesa, rvb, size * count * 4, 4);
		rvb->aos_offset = GET_START(rvb);
		rvb->aos_stride = size;
	}

	switch (size) {
	case 1:
		r300EmitVec4(ctx, rvb, data, stride, count);
		break;
	case 2:
		r300EmitVec8(ctx, rvb, data, stride, count);
		break;
	case 3:
		r300EmitVec12(ctx, rvb, data, stride, count);
		break;
	case 4:
		r300EmitVec16(ctx, rvb, data, stride, count);
		break;
	default:
		assert(0);
		break;
	}
}

#define DW_SIZE(x) ((inputs[tab[(x)]] << R300_DST_VEC_LOC_SHIFT) |	\
		    (attribptr[tab[(x)]]->size - 1) << R300_DATA_TYPE_0_SHIFT)

GLuint r300VAPInputRoute0(uint32_t * dst, GLvector4f ** attribptr,
				 int *inputs, GLint * tab, GLuint nr)
{
	GLuint i, dw;

	/* type, inputs, stop bit, size */
	for (i = 0; i < nr; i += 2) {
		/* make sure input is valid, would lockup the gpu */
		assert(inputs[tab[i]] != -1);
		dw = (R300_SIGNED | DW_SIZE(i));
		if (i + 1 == nr) {
			dw |= R300_LAST_VEC << R300_DATA_TYPE_0_SHIFT;
		} else {
			assert(inputs[tab[i + 1]] != -1);
			dw |= (R300_SIGNED |
			       DW_SIZE(i + 1)) << R300_DATA_TYPE_1_SHIFT;
			if (i + 2 == nr) {
				dw |= R300_LAST_VEC << R300_DATA_TYPE_1_SHIFT;
			}
		}
		dst[i >> 1] = dw;
	}

	return (nr + 1) >> 1;
}

static GLuint r300VAPInputRoute1Swizzle(int swizzle[4])
{
	return (swizzle[0] << R300_SWIZZLE_SELECT_X_SHIFT) |
	    (swizzle[1] << R300_SWIZZLE_SELECT_Y_SHIFT) |
	    (swizzle[2] << R300_SWIZZLE_SELECT_Z_SHIFT) |
	    (swizzle[3] << R300_SWIZZLE_SELECT_W_SHIFT);
}

GLuint r300VAPInputRoute1(uint32_t * dst, int swizzle[][4], GLuint nr)
{
	GLuint i, dw;

	for (i = 0; i < nr; i += 2) {
		dw = (r300VAPInputRoute1Swizzle(swizzle[i]) |
		      ((R300_WRITE_ENA_X | R300_WRITE_ENA_Y |
			R300_WRITE_ENA_Z | R300_WRITE_ENA_W) << R300_WRITE_ENA_SHIFT)) << R300_SWIZZLE0_SHIFT;
		if (i + 1 < nr) {
			dw |= (r300VAPInputRoute1Swizzle(swizzle[i + 1]) |
			       ((R300_WRITE_ENA_X | R300_WRITE_ENA_Y |
				 R300_WRITE_ENA_Z | R300_WRITE_ENA_W) << R300_WRITE_ENA_SHIFT)) << R300_SWIZZLE1_SHIFT;
		}
		dst[i >> 1] = dw;
	}

	return (nr + 1) >> 1;
}

GLuint r300VAPInputCntl0(GLcontext * ctx, GLuint InputsRead)
{
	/* No idea what this value means. I have seen other values written to
	 * this register... */
	return 0x5555;
}

GLuint r300VAPInputCntl1(GLcontext * ctx, GLuint InputsRead)
{
	r300ContextPtr rmesa = R300_CONTEXT(ctx);
	GLuint i, vic_1 = 0;

	if (InputsRead & (1 << VERT_ATTRIB_POS))
		vic_1 |= R300_INPUT_CNTL_POS;