linux/drivers/gpu/drm/radeon/r300_cmdbuf.c

/* r300_cmdbuf.c -- Command buffer emission for R300 -*- linux-c -*-
 *
 * Copyright (C) The Weather Channel, Inc.  2002.
 * Copyright (C) 2004 Nicolai Haehnle.
 * 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
 * PRECISION INSIGHT 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.
 *
 * Authors:
 *    Nicolai Haehnle <prefect_@gmx.net>
 */

#include "drmP.h"
#include "drm.h"
#include "drm_buffer.h"
#include "radeon_drm.h"
#include "radeon_drv.h"
#include "r300_reg.h"

#include <asm/unaligned.h>

#define R300_SIMULTANEOUS_CLIPRECTS		4

/* Values for R300_RE_CLIPRECT_CNTL depending on the number of cliprects
 */
static const int r300_cliprect_cntl[4] = {
	0xAAAA,
	0xEEEE,
	0xFEFE,
	0xFFFE
};

/**
 * Emit up to R300_SIMULTANEOUS_CLIPRECTS cliprects from the given command
 * buffer, starting with index n.
 */
static int r300_emit_cliprects(drm_radeon_private_t *dev_priv,
			       drm_radeon_kcmd_buffer_t *cmdbuf, int n)
{
	struct drm_clip_rect box;
	int nr;
	int i;
	RING_LOCALS;

	nr = cmdbuf->nbox - n;
	if (nr > R300_SIMULTANEOUS_CLIPRECTS)
		nr = R300_SIMULTANEOUS_CLIPRECTS;

	DRM_DEBUG("%i cliprects\n", nr);

	if (nr) {
		BEGIN_RING(6 + nr * 2);
		OUT_RING(CP_PACKET0(R300_RE_CLIPRECT_TL_0, nr * 2 - 1));

		for (i = 0; i < nr; ++i) {
			if (DRM_COPY_FROM_USER_UNCHECKED
			    (&box, &cmdbuf->boxes[n + i], sizeof(box))) {
				DRM_ERROR("copy cliprect faulted\n");
				return -EFAULT;
			}

			box.x2--; /* Hardware expects inclusive bottom-right corner */
			box.y2--;

			if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_RV515) {
				box.x1 = (box.x1) &
					R300_CLIPRECT_MASK;
				box.y1 = (box.y1) &
					R300_CLIPRECT_MASK;
				box.x2 = (box.x2) &
					R300_CLIPRECT_MASK;
				box.y2 = (box.y2) &
					R300_CLIPRECT_MASK;
			} else {
				box.x1 = (box.x1 + R300_CLIPRECT_OFFSET) &
					R300_CLIPRECT_MASK;
				box.y1 = (box.y1 + R300_CLIPRECT_OFFSET) &
					R300_CLIPRECT_MASK;
				box.x2 = (box.x2 + R300_CLIPRECT_OFFSET) &
					R300_CLIPRECT_MASK;
				box.y2 = (box.y2 + R300_CLIPRECT_OFFSET) &
					R300_CLIPRECT_MASK;
			}

			OUT_RING((box.x1 << R300_CLIPRECT_X_SHIFT) |
				 (box.y1 << R300_CLIPRECT_Y_SHIFT));
			OUT_RING((box.x2 << R300_CLIPRECT_X_SHIFT) |
				 (box.y2 << R300_CLIPRECT_Y_SHIFT));

		}

		OUT_RING_REG(R300_RE_CLIPRECT_CNTL, r300_cliprect_cntl[nr - 1]);

		/* TODO/SECURITY: Force scissors to a safe value, otherwise the
		 * client might be able to trample over memory.
		 * The impact should be very limited, but I'd rather be safe than
		 * sorry.
		 */
		OUT_RING(CP_PACKET0(R300_RE_SCISSORS_TL, 1));
		OUT_RING(0);
		OUT_RING(R300_SCISSORS_X_MASK | R300_SCISSORS_Y_MASK);
		ADVANCE_RING();
	} else {
		/* Why we allow zero cliprect rendering:
		 * There are some commands in a command buffer that must be submitted
		 * even when there are no cliprects, e.g. DMA buffer discard
		 * or state setting (though state setting could be avoided by
		 * simulating a loss of context).
		 *
		 * Now since the cmdbuf interface is so chaotic right now (and is
		 * bound to remain that way for a bit until things settle down),
		 * it is basically impossible to filter out the commands that are
		 * necessary and those that aren't.
		 *
		 * So I choose the safe way and don't do any filtering at all;
		 * instead, I simply set up the engine so that all rendering
		 * can't produce any fragments.
		 */
		BEGIN_RING(2);
		OUT_RING_REG(R300_RE_CLIPRECT_CNTL, 0);
		ADVANCE_RING();
	}

	/* flus cache and wait idle clean after cliprect change */
	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));
	OUT_RING(R300_RB3D_DC_FLUSH);
	ADVANCE_RING();
	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(RADEON_WAIT_UNTIL, 0));
	OUT_RING(RADEON_WAIT_3D_IDLECLEAN);
	ADVANCE_RING();
	/* set flush flag */
	dev_priv->track_flush |= RADEON_FLUSH_EMITED;

	return 0;
}

static u8 r300_reg_flags[0x10000 >> 2];

void r300_init_reg_flags(struct drm_device *dev)
{
	int i;
	drm_radeon_private_t *dev_priv = dev->dev_private;

	memset(r300_reg_flags, 0, 0x10000 >> 2);
#define ADD_RANGE_MARK(reg, count,mark) \
		for(i=((reg)>>2);i<((reg)>>2)+(count);i++)\
			r300_reg_flags[i]|=(mark);

#define MARK_SAFE		1
#define MARK_CHECK_OFFSET	2

#define ADD_RANGE(reg, count)	ADD_RANGE_MARK(reg, count, MARK_SAFE)

	/* these match cmducs() command in r300_driver/r300/r300_cmdbuf.c */
	ADD_RANGE(R300_SE_VPORT_XSCALE, 6);
	ADD_RANGE(R300_VAP_CNTL, 1);
	ADD_RANGE(R300_SE_VTE_CNTL, 2);
	ADD_RANGE(0x2134, 2);
	ADD_RANGE(R300_VAP_CNTL_STATUS, 1);
	ADD_RANGE(R300_VAP_INPUT_CNTL_0, 2);
	ADD_RANGE(0x21DC, 1);
	ADD_RANGE(R300_VAP_UNKNOWN_221C, 1);
	ADD_RANGE(R300_VAP_CLIP_X_0, 4);
	ADD_RANGE(R300_VAP_PVS_STATE_FLUSH_REG, 1);
	ADD_RANGE(R300_VAP_UNKNOWN_2288, 1);
	ADD_RANGE(R300_VAP_OUTPUT_VTX_FMT_0, 2);
	ADD_RANGE(R300_VAP_PVS_CNTL_1, 3);
	ADD_RANGE(R300_GB_ENABLE, 1);
	ADD_RANGE(R300_GB_MSPOS0, 5);
	ADD_RANGE(R300_TX_INVALTAGS, 1);
	ADD_RANGE(R300_TX_ENABLE, 1);
	ADD_RANGE(0x4200, 4);
	ADD_RANGE(0x4214, 1);
	ADD_RANGE(R300_RE_POINTSIZE, 1);
	ADD_RANGE(0x4230, 3);
	ADD_RANGE(R300_RE_LINE_CNT, 1);
	ADD_RANGE(R300_RE_UNK4238, 1);
	ADD_RANGE(0x4260, 3);
	ADD_RANGE(R300_RE_SHADE, 4);
	ADD_RANGE(R300_RE_POLYGON_MODE, 5);
	ADD_RANGE(R300_RE_ZBIAS_CNTL, 1);
	ADD_RANGE(R300_RE_ZBIAS_T_FACTOR, 4);
	ADD_RANGE(R300_RE_OCCLUSION_CNTL, 1);
	ADD_RANGE(R300_RE_CULL_CNTL, 1);
	ADD_RANGE(0x42C0, 2);
	ADD_RANGE(R300_RS_CNTL_0, 2);

	ADD_RANGE(R300_SU_REG_DEST, 1);
	if ((dev_priv->flags & RADEON_FAMILY_MASK) == CHIP_RV530)
		ADD_RANGE(RV530_FG_ZBREG_DEST, 1);

	ADD_RANGE(R300_SC_HYPERZ, 2);
	ADD_RANGE(0x43E8, 1);

	ADD_RANGE(0x46A4, 5);

	ADD_RANGE(R300_RE_FOG_STATE, 1);
	ADD_RANGE(R300_FOG_COLOR_R, 3);
	ADD_RANGE(R300_PP_ALPHA_TEST, 2);
	ADD_RANGE(0x4BD8, 1);
	ADD_RANGE(R300_PFS_PARAM_0_X, 64);
	ADD_RANGE(0x4E00, 1);
	ADD_RANGE(R300_RB3D_CBLEND, 2);
	ADD_RANGE(R300_RB3D_COLORMASK, 1);
	ADD_RANGE(R300_RB3D_BLEND_COLOR, 3);
	ADD_RANGE_MARK(R300_RB3D_COLOROFFSET0, 1, MARK_CHECK_OFFSET);	/* check offset */
	ADD_RANGE(R300_RB3D_COLORPITCH0, 1);
	ADD_RANGE(0x4E50, 9);
	ADD_RANGE(0x4E88, 1);
	ADD_RANGE(0x4EA0, 2);
	ADD_RANGE(R300_ZB_CNTL, 3);
	ADD_RANGE(R300_ZB_FORMAT, 4);
	ADD_RANGE_MARK(R300_ZB_DEPTHOFFSET, 1, MARK_CHECK_OFFSET);	/* check offset */
	ADD_RANGE(R300_ZB_DEPTHPITCH, 1);
	ADD_RANGE(R300_ZB_DEPTHCLEARVALUE, 1);
	ADD_RANGE(R300_ZB_ZMASK_OFFSET, 13);
	ADD_RANGE(R300_ZB_ZPASS_DATA, 2); /* ZB_ZPASS_DATA, ZB_ZPASS_ADDR */

	ADD_RANGE(R300_TX_FILTER_0, 16);
	ADD_RANGE(R300_TX_FILTER1_0, 16);
	ADD_RANGE(R300_TX_SIZE_0, 16);
	ADD_RANGE(R300_TX_FORMAT_0, 16);
	ADD_RANGE(R300_TX_PITCH_0, 16);
	/* Texture offset is dangerous and needs more checking */
	ADD_RANGE_MARK(R300_TX_OFFSET_0, 16, MARK_CHECK_OFFSET);
	ADD_RANGE(R300_TX_CHROMA_KEY_0, 16);
	ADD_RANGE(R300_TX_BORDER_COLOR_0, 16);

	/* Sporadic registers used as primitives are emitted */
	ADD_RANGE(R300_ZB_ZCACHE_CTLSTAT, 1);
	ADD_RANGE(R300_RB3D_DSTCACHE_CTLSTAT, 1);
	ADD_RANGE(R300_VAP_INPUT_ROUTE_0_0, 8);
	ADD_RANGE(R300_VAP_INPUT_ROUTE_1_0, 8);

	if ((dev_priv->flags & RADEON_FAMILY_MASK) >= CHIP_RV515) {
		ADD_RANGE(R500_VAP_INDEX_OFFSET, 1);
		ADD_RANGE(R500_US_CONFIG, 2);
		ADD_RANGE(R500_US_CODE_ADDR, 3);
		ADD_RANGE(R500_US_FC_CTRL, 1);
		ADD_RANGE(R500_RS_IP_0, 16);
		ADD_RANGE(R500_RS_INST_0, 16);
		ADD_RANGE(R500_RB3D_COLOR_CLEAR_VALUE_AR, 2);
		ADD_RANGE(R500_RB3D_CONSTANT_COLOR_AR, 2);
		ADD_RANGE(R500_ZB_FIFO_SIZE, 2);
	} else {
		ADD_RANGE(R300_PFS_CNTL_0, 3);
		ADD_RANGE(R300_PFS_NODE_0, 4);
		ADD_RANGE(R300_PFS_TEXI_0, 64);
		ADD_RANGE(R300_PFS_INSTR0_0, 64);
		ADD_RANGE(R300_PFS_INSTR1_0, 64);
		ADD_RANGE(R300_PFS_INSTR2_0, 64);
		ADD_RANGE(R300_PFS_INSTR3_0, 64);
		ADD_RANGE(R300_RS_INTERP_0, 8);
		ADD_RANGE(R300_RS_ROUTE_0, 8);

	}
}

static __inline__ int r300_check_range(unsigned reg, int count)
{
	int i;
	if (reg & ~0xffff)
		return -1;
	for (i = (reg >> 2); i < (reg >> 2) + count; i++)
		if (r300_reg_flags[i] != MARK_SAFE)
			return 1;
	return 0;
}

static __inline__ int r300_emit_carefully_checked_packet0(drm_radeon_private_t *
							  dev_priv,
							  drm_radeon_kcmd_buffer_t
							  * cmdbuf,
							  drm_r300_cmd_header_t
							  header)
{
	int reg;
	int sz;
	int i;
	u32 *value;
	RING_LOCALS;

	sz = header.packet0.count;
	reg = (header.packet0.reghi << 8) | header.packet0.reglo;

	if ((sz > 64) || (sz < 0)) {
		DRM_ERROR("Cannot emit more than 64 values at a time (reg=%04x sz=%d)\n",
			 reg, sz);
		return -EINVAL;
	}

	for (i = 0; i < sz; i++) {
		switch (r300_reg_flags[(reg >> 2) + i]) {
		case MARK_SAFE:
			break;
		case MARK_CHECK_OFFSET:
			value = drm_buffer_pointer_to_dword(cmdbuf->buffer, i);
			if (!radeon_check_offset(dev_priv, *value)) {
				DRM_ERROR("Offset failed range check (reg=%04x sz=%d)\n",
					 reg, sz);
				return -EINVAL;
			}
			break;
		default:
			DRM_ERROR("Register %04x failed check as flag=%02x\n",
				reg + i * 4, r300_reg_flags[(reg >> 2) + i]);
			return -EINVAL;
		}
	}

	BEGIN_RING(1 + sz);
	OUT_RING(CP_PACKET0(reg, sz - 1));
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, sz);
	ADVANCE_RING();

	return 0;
}

/**
 * Emits a packet0 setting arbitrary registers.
 * Called by r300_do_cp_cmdbuf.
 *
 * Note that checks are performed on contents and addresses of the registers
 */
static __inline__ int r300_emit_packet0(drm_radeon_private_t *dev_priv,
					drm_radeon_kcmd_buffer_t *cmdbuf,
					drm_r300_cmd_header_t header)
{
	int reg;
	int sz;
	RING_LOCALS;

	sz = header.packet0.count;
	reg = (header.packet0.reghi << 8) | header.packet0.reglo;

	if (!sz)
		return 0;

	if (sz * 4 > drm_buffer_unprocessed(cmdbuf->buffer))
		return -EINVAL;

	if (reg + sz * 4 >= 0x10000) {
		DRM_ERROR("No such registers in hardware reg=%04x sz=%d\n", reg,
			  sz);
		return -EINVAL;
	}

	if (r300_check_range(reg, sz)) {
		/* go and check everything */
		return r300_emit_carefully_checked_packet0(dev_priv, cmdbuf,
							   header);
	}
	/* the rest of the data is safe to emit, whatever the values the user passed */

	BEGIN_RING(1 + sz);
	OUT_RING(CP_PACKET0(reg, sz - 1));
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, sz);
	ADVANCE_RING();

	return 0;
}

/**
 * Uploads user-supplied vertex program instructions or parameters onto
 * the graphics card.
 * Called by r300_do_cp_cmdbuf.
 */
static __inline__ int r300_emit_vpu(drm_radeon_private_t *dev_priv,
				    drm_radeon_kcmd_buffer_t *cmdbuf,
				    drm_r300_cmd_header_t header)
{
	int sz;
	int addr;
	RING_LOCALS;

	sz = header.vpu.count;
	addr = (header.vpu.adrhi << 8) | header.vpu.adrlo;

	if (!sz)
		return 0;
	if (sz * 16 > drm_buffer_unprocessed(cmdbuf->buffer))
		return -EINVAL;

	/* VAP is very sensitive so we purge cache before we program it
	 * and we also flush its state before & after */
	BEGIN_RING(6);
	OUT_RING(CP_PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));
	OUT_RING(R300_RB3D_DC_FLUSH);
	OUT_RING(CP_PACKET0(RADEON_WAIT_UNTIL, 0));
	OUT_RING(RADEON_WAIT_3D_IDLECLEAN);
	OUT_RING(CP_PACKET0(R300_VAP_PVS_STATE_FLUSH_REG, 0));
	OUT_RING(0);
	ADVANCE_RING();
	/* set flush flag */
	dev_priv->track_flush |= RADEON_FLUSH_EMITED;

	BEGIN_RING(3 + sz * 4);
	OUT_RING_REG(R300_VAP_PVS_UPLOAD_ADDRESS, addr);
	OUT_RING(CP_PACKET0_TABLE(R300_VAP_PVS_UPLOAD_DATA, sz * 4 - 1));
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, sz * 4);
	ADVANCE_RING();

	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(R300_VAP_PVS_STATE_FLUSH_REG, 0));
	OUT_RING(0);
	ADVANCE_RING();

	return 0;
}

/**
 * Emit a clear packet from userspace.
 * Called by r300_emit_packet3.
 */
static __inline__ int r300_emit_clear(drm_radeon_private_t *dev_priv,
				      drm_radeon_kcmd_buffer_t *cmdbuf)
{
	RING_LOCALS;

	if (8 * 4 > drm_buffer_unprocessed(cmdbuf->buffer))
		return -EINVAL;

	BEGIN_RING(10);
	OUT_RING(CP_PACKET3(R200_3D_DRAW_IMMD_2, 8));
	OUT_RING(R300_PRIM_TYPE_POINT | R300_PRIM_WALK_RING |
		 (1 << R300_PRIM_NUM_VERTICES_SHIFT));
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, 8);
	ADVANCE_RING();

	BEGIN_RING(4);
	OUT_RING(CP_PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));
	OUT_RING(R300_RB3D_DC_FLUSH);
	OUT_RING(CP_PACKET0(RADEON_WAIT_UNTIL, 0));
	OUT_RING(RADEON_WAIT_3D_IDLECLEAN);
	ADVANCE_RING();
	/* set flush flag */
	dev_priv->track_flush |= RADEON_FLUSH_EMITED;

	return 0;
}

static __inline__ int r300_emit_3d_load_vbpntr(drm_radeon_private_t *dev_priv,
					       drm_radeon_kcmd_buffer_t *cmdbuf,
					       u32 header)
{
	int count, i, k;
#define MAX_ARRAY_PACKET  64
	u32 *data;
	u32 narrays;
	RING_LOCALS;

	count = (header & RADEON_CP_PACKET_COUNT_MASK) >> 16;

	if ((count + 1) > MAX_ARRAY_PACKET) {
		DRM_ERROR("Too large payload in 3D_LOAD_VBPNTR (count=%d)\n",
			  count);
		return -EINVAL;
	}
	/* carefully check packet contents */

	/* We have already read the header so advance the buffer. */
	drm_buffer_advance(cmdbuf->buffer, 4);

	narrays = *(u32 *)drm_buffer_pointer_to_dword(cmdbuf->buffer, 0);
	k = 0;
	i = 1;
	while ((k < narrays) && (i < (count + 1))) {
		i++;		/* skip attribute field */
		data = drm_buffer_pointer_to_dword(cmdbuf->buffer, i);
		if (!radeon_check_offset(dev_priv, *data)) {
			DRM_ERROR
			    ("Offset failed range check (k=%d i=%d) while processing 3D_LOAD_VBPNTR packet.\n",
			     k, i);
			return -EINVAL;
		}
		k++;
		i++;
		if (k == narrays)
			break;
		/* have one more to process, they come in pairs */
		data = drm_buffer_pointer_to_dword(cmdbuf->buffer, i);
		if (!radeon_check_offset(dev_priv, *data)) {
			DRM_ERROR
			    ("Offset failed range check (k=%d i=%d) while processing 3D_LOAD_VBPNTR packet.\n",
			     k, i);
			return -EINVAL;
		}
		k++;
		i++;
	}
	/* do the counts match what we expect ? */
	if ((k != narrays) || (i != (count + 1))) {
		DRM_ERROR
		    ("Malformed 3D_LOAD_VBPNTR packet (k=%d i=%d narrays=%d count+1=%d).\n",
		     k, i, narrays, count + 1);
		return -EINVAL;
	}

	/* all clear, output packet */

	BEGIN_RING(count + 2);
	OUT_RING(header);
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, count + 1);
	ADVANCE_RING();

	return 0;
}

static __inline__ int r300_emit_bitblt_multi(drm_radeon_private_t *dev_priv,
					     drm_radeon_kcmd_buffer_t *cmdbuf)
{
	u32 *cmd = drm_buffer_pointer_to_dword(cmdbuf->buffer, 0);
	int count, ret;
	RING_LOCALS;


	count = (*cmd & RADEON_CP_PACKET_COUNT_MASK) >> 16;

	if (*cmd & 0x8000) {
		u32 offset;
		u32 *cmd1 = drm_buffer_pointer_to_dword(cmdbuf->buffer, 1);
		if (*cmd1 & (RADEON_GMC_SRC_PITCH_OFFSET_CNTL
			      | RADEON_GMC_DST_PITCH_OFFSET_CNTL)) {

			u32 *cmd2 = drm_buffer_pointer_to_dword(cmdbuf->buffer, 2);
			offset = *cmd2 << 10;
			ret = !radeon_check_offset(dev_priv, offset);
			if (ret) {
				DRM_ERROR("Invalid bitblt first offset is %08X\n", offset);
				return -EINVAL;
			}
		}

		if ((*cmd1 & RADEON_GMC_SRC_PITCH_OFFSET_CNTL) &&
		    (*cmd1 & RADEON_GMC_DST_PITCH_OFFSET_CNTL)) {
			u32 *cmd3 = drm_buffer_pointer_to_dword(cmdbuf->buffer, 3);
			offset = *cmd3 << 10;
			ret = !radeon_check_offset(dev_priv, offset);
			if (ret) {
				DRM_ERROR("Invalid bitblt second offset is %08X\n", offset);
				return -EINVAL;
			}

		}
	}

	BEGIN_RING(count+2);
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, count + 2);
	ADVANCE_RING();

	return 0;
}

static __inline__ int r300_emit_draw_indx_2(drm_radeon_private_t *dev_priv,
					    drm_radeon_kcmd_buffer_t *cmdbuf)
{
	u32 *cmd = drm_buffer_pointer_to_dword(cmdbuf->buffer, 0);
	u32 *cmd1 = drm_buffer_pointer_to_dword(cmdbuf->buffer, 1);
	int count;
	int expected_count;
	RING_LOCALS;

	count = (*cmd & RADEON_CP_PACKET_COUNT_MASK) >> 16;

	expected_count = *cmd1 >> 16;
	if (!(*cmd1 & R300_VAP_VF_CNTL__INDEX_SIZE_32bit))
		expected_count = (expected_count+1)/2;

	if (count && count != expected_count) {
		DRM_ERROR("3D_DRAW_INDX_2: packet size %i, expected %i\n",
			count, expected_count);
		return -EINVAL;
	}

	BEGIN_RING(count+2);
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, count + 2);
	ADVANCE_RING();

	if (!count) {
		drm_r300_cmd_header_t stack_header, *header;
		u32 *cmd1, *cmd2, *cmd3;

		if (drm_buffer_unprocessed(cmdbuf->buffer)
				< 4*4 + sizeof(stack_header)) {
			DRM_ERROR("3D_DRAW_INDX_2: expect subsequent INDX_BUFFER, but stream is too short.\n");
			return -EINVAL;
		}

		header = drm_buffer_read_object(cmdbuf->buffer,
				sizeof(stack_header), &stack_header);

		cmd = drm_buffer_pointer_to_dword(cmdbuf->buffer, 0);
		cmd1 = drm_buffer_pointer_to_dword(cmdbuf->buffer, 1);
		cmd2 = drm_buffer_pointer_to_dword(cmdbuf->buffer, 2);
		cmd3 = drm_buffer_pointer_to_dword(cmdbuf->buffer, 3);

		if (header->header.cmd_type != R300_CMD_PACKET3 ||
		    header->packet3.packet != R300_CMD_PACKET3_RAW ||
		    *cmd != CP_PACKET3(RADEON_CP_INDX_BUFFER, 2)) {
			DRM_ERROR("3D_DRAW_INDX_2: expect subsequent INDX_BUFFER.\n");
			return -EINVAL;
		}

		if ((*cmd1 & 0x8000ffff) != 0x80000810) {
			DRM_ERROR("Invalid indx_buffer reg address %08X\n",
					*cmd1);
			return -EINVAL;
		}
		if (!radeon_check_offset(dev_priv, *cmd2)) {
			DRM_ERROR("Invalid indx_buffer offset is %08X\n",
					*cmd2);
			return -EINVAL;
		}
		if (*cmd3 != expected_count) {
			DRM_ERROR("INDX_BUFFER: buffer size %i, expected %i\n",
				*cmd3, expected_count);
			return -EINVAL;
		}

		BEGIN_RING(4);
		OUT_RING_DRM_BUFFER(cmdbuf->buffer, 4);
		ADVANCE_RING();
	}

	return 0;
}

static __inline__ int r300_emit_raw_packet3(drm_radeon_private_t *dev_priv,
					    drm_radeon_kcmd_buffer_t *cmdbuf)
{
	u32 *header;
	int count;
	RING_LOCALS;

	if (4 > drm_buffer_unprocessed(cmdbuf->buffer))
		return -EINVAL;

	/* Fixme !! This simply emits a packet without much checking.
	   We need to be smarter. */

	/* obtain first word - actual packet3 header */
	header = drm_buffer_pointer_to_dword(cmdbuf->buffer, 0);

	/* Is it packet 3 ? */
	if ((*header >> 30) != 0x3) {
		DRM_ERROR("Not a packet3 header (0x%08x)\n", *header);
		return -EINVAL;
	}

	count = (*header >> 16) & 0x3fff;

	/* Check again now that we know how much data to expect */
	if ((count + 2) * 4 > drm_buffer_unprocessed(cmdbuf->buffer)) {
		DRM_ERROR
		    ("Expected packet3 of length %d but have only %d bytes left\n",
		     (count + 2) * 4, drm_buffer_unprocessed(cmdbuf->buffer));
		return -EINVAL;
	}

	/* Is it a packet type we know about ? */
	switch (*header & 0xff00) {
	case RADEON_3D_LOAD_VBPNTR:	/* load vertex array pointers */
		return r300_emit_3d_load_vbpntr(dev_priv, cmdbuf, *header);

	case RADEON_CNTL_BITBLT_MULTI:
		return r300_emit_bitblt_multi(dev_priv, cmdbuf);

	case RADEON_CP_INDX_BUFFER:
		DRM_ERROR("packet3 INDX_BUFFER without preceding 3D_DRAW_INDX_2 is illegal.\n");
		return -EINVAL;
	case RADEON_CP_3D_DRAW_IMMD_2:
		/* triggers drawing using in-packet vertex data */
	case RADEON_CP_3D_DRAW_VBUF_2:
		/* triggers drawing of vertex buffers setup elsewhere */
		dev_priv->track_flush &= ~(RADEON_FLUSH_EMITED |
					   RADEON_PURGE_EMITED);
		break;
	case RADEON_CP_3D_DRAW_INDX_2:
		/* triggers drawing using indices to vertex buffer */
		/* whenever we send vertex we clear flush & purge */
		dev_priv->track_flush &= ~(RADEON_FLUSH_EMITED |
					   RADEON_PURGE_EMITED);
		return r300_emit_draw_indx_2(dev_priv, cmdbuf);
	case RADEON_WAIT_FOR_IDLE:
	case RADEON_CP_NOP:
		/* these packets are safe */
		break;
	default:
		DRM_ERROR("Unknown packet3 header (0x%08x)\n", *header);
		return -EINVAL;
	}

	BEGIN_RING(count + 2);
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, count + 2);
	ADVANCE_RING();

	return 0;
}

/**
 * Emit a rendering packet3 from userspace.
 * Called by r300_do_cp_cmdbuf.
 */
static __inline__ int r300_emit_packet3(drm_radeon_private_t *dev_priv,
					drm_radeon_kcmd_buffer_t *cmdbuf,
					drm_r300_cmd_header_t header)
{
	int n;
	int ret;
	int orig_iter = cmdbuf->buffer->iterator;

	/* This is a do-while-loop so that we run the interior at least once,
	 * even if cmdbuf->nbox is 0. Compare r300_emit_cliprects for rationale.
	 */
	n = 0;
	do {
		if (cmdbuf->nbox > R300_SIMULTANEOUS_CLIPRECTS) {
			ret = r300_emit_cliprects(dev_priv, cmdbuf, n);
			if (ret)
				return ret;

			cmdbuf->buffer->iterator = orig_iter;
		}

		switch (header.packet3.packet) {
		case R300_CMD_PACKET3_CLEAR:
			DRM_DEBUG("R300_CMD_PACKET3_CLEAR\n");
			ret = r300_emit_clear(dev_priv, cmdbuf);
			if (ret) {
				DRM_ERROR("r300_emit_clear failed\n");
				return ret;
			}
			break;

		case R300_CMD_PACKET3_RAW:
			DRM_DEBUG("R300_CMD_PACKET3_RAW\n");
			ret = r300_emit_raw_packet3(dev_priv, cmdbuf);
			if (ret) {
				DRM_ERROR("r300_emit_raw_packet3 failed\n");
				return ret;
			}
			break;

		default:
			DRM_ERROR("bad packet3 type %i at byte %d\n",
				  header.packet3.packet,
				  cmdbuf->buffer->iterator - (int)sizeof(header));
			return -EINVAL;
		}

		n += R300_SIMULTANEOUS_CLIPRECTS;
	} while (n < cmdbuf->nbox);

	return 0;
}

/* Some of the R300 chips seem to be extremely touchy about the two registers
 * that are configured in r300_pacify.
 * Among the worst offenders seems to be the R300 ND (0x4E44): When userspace
 * sends a command buffer that contains only state setting commands and a
 * vertex program/parameter upload sequence, this will eventually lead to a
 * lockup, unless the sequence is bracketed by calls to r300_pacify.
 * So we should take great care to *always* call r300_pacify before
 * *anything* 3D related, and again afterwards. This is what the
 * call bracket in r300_do_cp_cmdbuf is for.
 */

/**
 * Emit the sequence to pacify R300.
 */
static void r300_pacify(drm_radeon_private_t *dev_priv)
{
	uint32_t cache_z, cache_3d, cache_2d;
	RING_LOCALS;

	cache_z = R300_ZC_FLUSH;
	cache_2d = R300_RB2D_DC_FLUSH;
	cache_3d = R300_RB3D_DC_FLUSH;
	if (!(dev_priv->track_flush & RADEON_PURGE_EMITED)) {
		/* we can purge, primitive where draw since last purge */
		cache_z |= R300_ZC_FREE;
		cache_2d |= R300_RB2D_DC_FREE;
		cache_3d |= R300_RB3D_DC_FREE;
	}

	/* flush & purge zbuffer */
	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(R300_ZB_ZCACHE_CTLSTAT, 0));
	OUT_RING(cache_z);
	ADVANCE_RING();
	/* flush & purge 3d */
	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(R300_RB3D_DSTCACHE_CTLSTAT, 0));
	OUT_RING(cache_3d);
	ADVANCE_RING();
	/* flush & purge texture */
	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(R300_TX_INVALTAGS, 0));
	OUT_RING(0);
	ADVANCE_RING();
	/* FIXME: is this one really needed ? */
	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(R300_RB3D_AARESOLVE_CTL, 0));
	OUT_RING(0);
	ADVANCE_RING();
	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(RADEON_WAIT_UNTIL, 0));
	OUT_RING(RADEON_WAIT_3D_IDLECLEAN);
	ADVANCE_RING();
	/* flush & purge 2d through E2 as RB2D will trigger lockup */
	BEGIN_RING(4);
	OUT_RING(CP_PACKET0(R300_DSTCACHE_CTLSTAT, 0));
	OUT_RING(cache_2d);
	OUT_RING(CP_PACKET0(RADEON_WAIT_UNTIL, 0));
	OUT_RING(RADEON_WAIT_2D_IDLECLEAN |
		 RADEON_WAIT_HOST_IDLECLEAN);
	ADVANCE_RING();
	/* set flush & purge flags */
	dev_priv->track_flush |= RADEON_FLUSH_EMITED | RADEON_PURGE_EMITED;
}

/**
 * Called by r300_do_cp_cmdbuf to update the internal buffer age and state.
 * The actual age emit is done by r300_do_cp_cmdbuf, which is why you must
 * be careful about how this function is called.
 */
static void r300_discard_buffer(struct drm_device *dev, struct drm_master *master, struct drm_buf *buf)
{
	drm_radeon_buf_priv_t *buf_priv = buf->dev_private;
	struct drm_radeon_master_private *master_priv = master->driver_priv;

	buf_priv->age = ++master_priv->sarea_priv->last_dispatch;
	buf->pending = 1;
	buf->used = 0;
}

static void r300_cmd_wait(drm_radeon_private_t * dev_priv,
			  drm_r300_cmd_header_t header)
{
	u32 wait_until;
	RING_LOCALS;

	if (!header.wait.flags)
		return;

	wait_until = 0;

	switch(header.wait.flags) {
	case R300_WAIT_2D:
		wait_until = RADEON_WAIT_2D_IDLE;
		break;
	case R300_WAIT_3D:
		wait_until = RADEON_WAIT_3D_IDLE;
		break;
	case R300_NEW_WAIT_2D_3D:
		wait_until = RADEON_WAIT_2D_IDLE|RADEON_WAIT_3D_IDLE;
		break;
	case R300_NEW_WAIT_2D_2D_CLEAN:
		wait_until = RADEON_WAIT_2D_IDLE|RADEON_WAIT_2D_IDLECLEAN;
		break;
	case R300_NEW_WAIT_3D_3D_CLEAN:
		wait_until = RADEON_WAIT_3D_IDLE|RADEON_WAIT_3D_IDLECLEAN;
		break;
	case R300_NEW_WAIT_2D_2D_CLEAN_3D_3D_CLEAN:
		wait_until = RADEON_WAIT_2D_IDLE|RADEON_WAIT_2D_IDLECLEAN;
		wait_until |= RADEON_WAIT_3D_IDLE|RADEON_WAIT_3D_IDLECLEAN;
		break;
	default:
		return;
	}

	BEGIN_RING(2);
	OUT_RING(CP_PACKET0(RADEON_WAIT_UNTIL, 0));
	OUT_RING(wait_until);
	ADVANCE_RING();
}

static int r300_scratch(drm_radeon_private_t *dev_priv,
			drm_radeon_kcmd_buffer_t *cmdbuf,
			drm_r300_cmd_header_t header)
{
	u32 *ref_age_base;
	u32 i, *buf_idx, h_pending;
	u64 *ptr_addr;
	u64 stack_ptr_addr;
	RING_LOCALS;

	if (drm_buffer_unprocessed(cmdbuf->buffer) <
	    (sizeof(u64) + header.scratch.n_bufs * sizeof(*buf_idx))) {
		return -EINVAL;
	}

	if (header.scratch.reg >= 5) {
		return -EINVAL;
	}

	dev_priv->scratch_ages[header.scratch.reg]++;

	ptr_addr = drm_buffer_read_object(cmdbuf->buffer,
			sizeof(stack_ptr_addr), &stack_ptr_addr);
	ref_age_base = (u32 *)(unsigned long)get_unaligned(ptr_addr);

	for (i=0; i < header.scratch.n_bufs; i++) {
		buf_idx = drm_buffer_pointer_to_dword(cmdbuf->buffer, 0);
		*buf_idx *= 2; /* 8 bytes per buf */

		if (DRM_COPY_TO_USER(ref_age_base + *buf_idx,
				&dev_priv->scratch_ages[header.scratch.reg],
				sizeof(u32)))
			return -EINVAL;

		if (DRM_COPY_FROM_USER(&h_pending,
				ref_age_base + *buf_idx + 1,
				sizeof(u32)))
			return -EINVAL;

		if (h_pending == 0)
			return -EINVAL;

		h_pending--;

		if (DRM_COPY_TO_USER(ref_age_base + *buf_idx + 1,
					&h_pending,
					sizeof(u32)))
			return -EINVAL;

		drm_buffer_advance(cmdbuf->buffer, sizeof(*buf_idx));
	}

	BEGIN_RING(2);
	OUT_RING( CP_PACKET0( RADEON_SCRATCH_REG0 + header.scratch.reg * 4, 0 ) );
	OUT_RING( dev_priv->scratch_ages[header.scratch.reg] );
	ADVANCE_RING();

	return 0;
}

/**
 * Uploads user-supplied vertex program instructions or parameters onto
 * the graphics card.
 * Called by r300_do_cp_cmdbuf.
 */
static inline int r300_emit_r500fp(drm_radeon_private_t *dev_priv,
				       drm_radeon_kcmd_buffer_t *cmdbuf,
				       drm_r300_cmd_header_t header)
{
	int sz;
	int addr;
	int type;
	int isclamp;
	int stride;
	RING_LOCALS;

	sz = header.r500fp.count;
	/* address is 9 bits 0 - 8, bit 1 of flags is part of address */
	addr = ((header.r500fp.adrhi_flags & 1) << 8) | header.r500fp.adrlo;

	type = !!(header.r500fp.adrhi_flags & R500FP_CONSTANT_TYPE);
	isclamp = !!(header.r500fp.adrhi_flags & R500FP_CONSTANT_CLAMP);

	addr |= (type << 16);
	addr |= (isclamp << 17);

	stride = type ? 4 : 6;

	DRM_DEBUG("r500fp %d %d type: %d\n", sz, addr, type);
	if (!sz)
		return 0;
	if (sz * stride * 4 > drm_buffer_unprocessed(cmdbuf->buffer))
		return -EINVAL;

	BEGIN_RING(3 + sz * stride);
	OUT_RING_REG(R500_GA_US_VECTOR_INDEX, addr);
	OUT_RING(CP_PACKET0_TABLE(R500_GA_US_VECTOR_DATA, sz * stride - 1));
	OUT_RING_DRM_BUFFER(cmdbuf->buffer, sz * stride);

	ADVANCE_RING();

	return 0;
}


/**
 * Parses and validates a user-supplied command buffer and emits appropriate
 * commands on the DMA ring buffer.
 * Called by the ioctl handler function radeon_cp_cmdbuf.
 */
int r300_do_cp_cmdbuf(struct drm_device *dev,
		      struct drm_file *file_priv,
		      drm_radeon_kcmd_buffer_t *cmdbuf)
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	struct drm_radeon_master_private *master_priv = file_priv->master->driver_priv;
	struct drm_device_dma *dma = dev->dma;
	struct drm_buf *buf = NULL;
	int emit_dispatch_age = 0;
	int ret = 0;

	DRM_DEBUG("\n");

	/* pacify */
	r300_pacify(dev_priv);

	if (cmdbuf->nbox <= R300_SIMULTANEOUS_CLIPRECTS) {
		ret = r300_emit_cliprects(dev_priv, cmdbuf, 0);
		if (ret)
			goto cleanup;
	}

	while (drm_buffer_unprocessed(cmdbuf->buffer)
			>= sizeof(drm_r300_cmd_header_t)) {
		int idx;
		drm_r300_cmd_header_t *header, stack_header;

		header = drm_buffer_read_object(cmdbuf->buffer,
				sizeof(stack_header), &stack_header);

		switch (header->header.cmd_type) {
		case R300_CMD_PACKET0:
			DRM_DEBUG("R300_CMD_PACKET0\n");
			ret = r300_emit_packet0(dev_priv, cmdbuf, *header);
			if (ret) {
				DRM_ERROR("r300_emit_packet0 failed\n");
				goto cleanup;
			}
			break;

		case R300_CMD_VPU:
			DRM_DEBUG("R300_CMD_VPU\n");
			ret = r300_emit_vpu(dev_priv, cmdbuf, *header);
			if (ret) {
				DRM_ERROR("r300_emit_vpu failed\n");
				goto cleanup;
			}
			break;

		case R300_CMD_PACKET3:
			DRM_DEBUG("R300_CMD_PACKET3\n");
			ret = r300_emit_packet3(dev_priv, cmdbuf, *header);
			if (ret) {
				DRM_ERROR("r300_emit_packet3 failed\n");
				goto cleanup;
			}
			break;

		case R300_CMD_END3D:
			DRM_DEBUG("R300_CMD_END3D\n");
			/* TODO:
			   Ideally userspace driver should not need to issue this call,
			   i.e. the drm driver should issue it automatically and prevent
			   lockups.

			   In practice, we do not understand why this call is needed and what
			   it does (except for some vague guesses that it has to do with cache
			   coherence) and so the user space driver does it.

			   Once we are sure which uses prevent lockups the code could be moved
			   into the kernel and the userspace driver will not
			   need to use this command.

			   Note that issuing this command does not hurt anything
			   except, possibly, performance */
			r300_pacify(dev_priv);
			break;

		case R300_CMD_CP_DELAY:
			/* simple enough, we can do it here */
			DRM_DEBUG("R300_CMD_CP_DELAY\n");
			{
				int i;
				RING_LOCALS;

				BEGIN_RING(header->delay.count);
				for (i = 0; i < header->delay.count; i++)
					OUT_RING(RADEON_CP_PACKET2);
				ADVANCE_RING();
			}
			break;

		case R300_CMD_DMA_DISCARD:
			DRM_DEBUG("RADEON_CMD_DMA_DISCARD\n");
			idx = header->dma.buf_idx;
			if (idx < 0 || idx >= dma->buf_count) {
				DRM_ERROR("buffer index %d (of %d max)\n",
					  idx, dma->buf_count - 1);
				ret = -EINVAL;
				goto cleanup;
			}

			buf = dma->buflist[idx];
			if (buf->file_priv != file_priv || buf->pending) {
				DRM_ERROR("bad buffer %p %p %d\n",
					  buf->file_priv, file_priv,
					  buf->pending);
				ret = -EINVAL;
				goto cleanup;
			}

			emit_dispatch_age = 1;
			r300_discard_buffer(dev, file_priv->master, buf);
			break;

		case R300_CMD_WAIT:
			DRM_DEBUG("R300_CMD_WAIT\n");
			r300_cmd_wait(dev_priv, *header);
			break;

		case R300_CMD_SCRATCH:
			DRM_DEBUG("R300_CMD_SCRATCH\n");
			ret = r300_scratch(dev_priv, cmdbuf, *header);
			if (ret) {
				DRM_ERROR("r300_scratch failed\n");
				goto cleanup;
			}
			break;

		case R300_CMD_R500FP:
			if ((dev_priv->flags & RADEON_FAMILY_MASK) < CHIP_RV515) {
				DRM_ERROR("Calling r500 command on r300 card\n");
				ret = -EINVAL;
				goto cleanup;
			}
			DRM_DEBUG("R300_CMD_R500FP\n");
			ret = r300_emit_r500fp(dev_priv, cmdbuf, *header);
			if (ret) {
				DRM_ERROR("r300_emit_r500fp failed\n");
				goto cleanup;
			}
			break;
		default:
			DRM_ERROR("bad cmd_type %i at byte %d\n",
				  header->header.cmd_type,
				  cmdbuf->buffer->iterator - (int)sizeof(*header));
			ret = -EINVAL;
			goto cleanup;
		}
	}

	DRM_DEBUG("END\n");

      cleanup:
	r300_pacify(dev_priv);

	/* We emit the vertex buffer age here, outside the pacifier "brackets"
	 * for two reasons:
	 *  (1) This may coalesce multiple age emissions into a single one and
	 *  (2) more importantly, some chips lock up hard when scratch registers
	 *      are written inside the pacifier bracket.
	 */
	if (emit_dispatch_age) {
		RING_LOCALS;

		/* Emit the vertex buffer age */
		BEGIN_RING(2);
		RADEON_DISPATCH_AGE(master_priv->sarea_priv->last_dispatch);
		ADVANCE_RING();
	}

	COMMIT_RING();

	return ret;
}