summaryrefslogtreecommitdiff
path: root/shared/sis_mm.c
blob: 27b904fc7e31dab82b56c08a34c38c72a1c674ee (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
/* sis_mm.c -- Private header for Direct Rendering Manager -*- linux-c -*-
 * Created: Mon Jan  4 10:05:05 1999 by sclin@sis.com.tw
 *
 * Copyright 2000 Silicon Integrated Systems Corp, Inc., HsinChu, Taiwan.
 * All rights reserved.
 *
 * 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:
 *    Sung-Ching Lin <sclin@sis.com.tw>
 * 
 */

#include "sis.h"
#include "drmP.h"
#include "sis_drm.h"
#include "sis_drv.h"
#include "sis_ds.h"
#if defined(__linux__) && defined(CONFIG_FB_SIS)
#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,0)
#include <video/sisfb.h>
#else
#include <linux/sisfb.h>
#endif
#endif

#define MAX_CONTEXT 100
#define VIDEO_TYPE 0 
#define AGP_TYPE 1

typedef struct {
	int used;
	int context;
	set_t *sets[2]; /* 0 for video, 1 for AGP */
} sis_context_t;

static sis_context_t global_ppriv[MAX_CONTEXT];


static int add_alloc_set(int context, int type, unsigned int val)
{
	int i, retval = 0;
	
	for (i = 0; i < MAX_CONTEXT; i++) {
		if (global_ppriv[i].used && global_ppriv[i].context == context)
		{
			retval = setAdd(global_ppriv[i].sets[type], val);
			break;
		}
	}
	return retval;
}

static int del_alloc_set(int context, int type, unsigned int val)
{  
	int i, retval = 0;

	for (i = 0; i < MAX_CONTEXT; i++) {
		if (global_ppriv[i].used && global_ppriv[i].context == context)
		{
			retval = setDel(global_ppriv[i].sets[type], val);
			break;
		}
	}
	return retval;
}

/* fb management via fb device */ 
#if defined(__linux__) && defined(CONFIG_FB_SIS)

int sis_fb_init( DRM_IOCTL_ARGS )
{
	return 0;
}

int sis_fb_alloc( DRM_IOCTL_ARGS )
{
	drm_sis_mem_t fb;
	struct sis_memreq req;
	drm_sis_mem_t __user *argp = (void __user *)data;
	int retval = 0;

	DRM_COPY_FROM_USER_IOCTL(fb, argp, sizeof(fb));

	req.size = fb.size;
	sis_malloc(&req);
	if (req.offset) {
		/* TODO */
		fb.offset = req.offset;
		fb.free = req.offset;
		if (!add_alloc_set(fb.context, VIDEO_TYPE, fb.free)) {
			DRM_DEBUG("adding to allocation set fails\n");
			sis_free(req.offset);
			retval = DRM_ERR(EINVAL);
		}
	} else {  
		fb.offset = 0;
		fb.size = 0;
		fb.free = 0;
	}

	DRM_COPY_TO_USER_IOCTL(argp, fb, sizeof(fb));

	DRM_DEBUG("alloc fb, size = %d, offset = %ld\n", fb.size, req.offset);

	return retval;
}

int sis_fb_free( DRM_IOCTL_ARGS )
{
	drm_sis_mem_t fb;
	int retval = 0;

	DRM_COPY_FROM_USER_IOCTL(fb, (drm_sis_mem_t __user *)data, sizeof(fb));

	if (!fb.free)
		return DRM_ERR(EINVAL);

	if (!del_alloc_set(fb.context, VIDEO_TYPE, fb.free))
		retval = DRM_ERR(EINVAL);
	sis_free(fb.free);

	DRM_DEBUG("free fb, offset = 0x%lx\n", fb.free);

	return retval;
}

#else

/* Called by the X Server to initialize the FB heap.  Allocations will fail
 * unless this is called.  Offset is the beginning of the heap from the
 * framebuffer offset (MaxXFBMem in XFree86).
 *
 * Memory layout according to Thomas Winischofer:
 * |------------------|DDDDDDDDDDDDDDDDDDDDDDDDDDDDD|HHHH|CCCCCCCCCCC|
 *
 *    X driver/sisfb                                  HW-   Command-
 *  framebuffer memory           DRI heap           Cursor   queue
 */
int sis_fb_init( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_sis_private_t *dev_priv = dev->dev_private;
	drm_sis_fb_t fb;

	DRM_COPY_FROM_USER_IOCTL(fb, (drm_sis_fb_t __user *)data, sizeof(fb));

	if (dev_priv == NULL) {
		dev->dev_private = DRM(calloc)(1, sizeof(drm_sis_private_t),
		    DRM_MEM_DRIVER);
		dev_priv = dev->dev_private;
		if (dev_priv == NULL)
			return ENOMEM;
	}

	if (dev_priv->FBHeap != NULL)
		return DRM_ERR(EINVAL);

	dev_priv->FBHeap = mmInit(fb.offset, fb.size);

	DRM_DEBUG("offset = %u, size = %u", fb.offset, fb.size);

	return 0;
}

int sis_fb_alloc( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_sis_private_t *dev_priv = dev->dev_private;
	drm_sis_mem_t __user *argp = (void __user *)data;
	drm_sis_mem_t fb;
	PMemBlock block;
	int retval = 0;

	if (dev_priv == NULL || dev_priv->FBHeap == NULL)
		return DRM_ERR(EINVAL);
  
	DRM_COPY_FROM_USER_IOCTL(fb, argp, sizeof(fb));
  
	block = mmAllocMem(dev_priv->FBHeap, fb.size, 0, 0);
	if (block) {
		/* TODO */
		fb.offset = block->ofs;
		fb.free = (unsigned long)block;
		if (!add_alloc_set(fb.context, VIDEO_TYPE, fb.free)) {
			DRM_DEBUG("adding to allocation set fails\n");
			mmFreeMem((PMemBlock)fb.free);
			retval = DRM_ERR(EINVAL);
		}
	} else {
		fb.offset = 0;
		fb.size = 0;
		fb.free = 0;
	}

	DRM_COPY_TO_USER_IOCTL(argp, fb, sizeof(fb));

	DRM_DEBUG("alloc fb, size = %d, offset = %d\n", fb.size, fb.offset);

	return retval;
}

int sis_fb_free( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_sis_private_t *dev_priv = dev->dev_private;
	drm_sis_mem_t fb;

	if (dev_priv == NULL || dev_priv->FBHeap == NULL)
		return DRM_ERR(EINVAL);

	DRM_COPY_FROM_USER_IOCTL(fb, (drm_sis_mem_t __user *)data, sizeof(fb));

	if (!mmBlockInHeap(dev_priv->FBHeap, (PMemBlock)fb.free))
		return DRM_ERR(EINVAL);

	if (!del_alloc_set(fb.context, VIDEO_TYPE, fb.free))
		return DRM_ERR(EINVAL);
	mmFreeMem((PMemBlock)fb.free);

	DRM_DEBUG("free fb, free = 0x%lx\n", fb.free);

	return 0;
}

#endif

/* agp memory management */ 

int sis_ioctl_agp_init( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_sis_private_t *dev_priv = dev->dev_private;
	drm_sis_agp_t agp;

	if (dev_priv == NULL) {
		dev->dev_private = DRM(calloc)(1, sizeof(drm_sis_private_t),
		    DRM_MEM_DRIVER);
		dev_priv = dev->dev_private;
		if (dev_priv == NULL)
			return ENOMEM;
	}

	if (dev_priv->AGPHeap != NULL)
		return DRM_ERR(EINVAL);

	DRM_COPY_FROM_USER_IOCTL(agp, (drm_sis_agp_t __user *)data, sizeof(agp));

	dev_priv->AGPHeap = mmInit(agp.offset, agp.size);

	DRM_DEBUG("offset = %u, size = %u", agp.offset, agp.size);
  
	return 0;
}

int sis_ioctl_agp_alloc( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_sis_private_t *dev_priv = dev->dev_private;
	drm_sis_mem_t __user *argp = (void __user *)data;
	drm_sis_mem_t agp;
	PMemBlock block;
	int retval = 0;
   
	if (dev_priv == NULL || dev_priv->AGPHeap == NULL)
		return DRM_ERR(EINVAL);
  
	DRM_COPY_FROM_USER_IOCTL(agp, argp, sizeof(agp));
  
	block = mmAllocMem(dev_priv->AGPHeap, agp.size, 0, 0);
	if (block) {
		/* TODO */
		agp.offset = block->ofs;
		agp.free = (unsigned long)block;
		if (!add_alloc_set(agp.context, AGP_TYPE, agp.free)) {
			DRM_DEBUG("adding to allocation set fails\n");
			mmFreeMem((PMemBlock)agp.free);
			retval = -1;
		}
	} else {  
		agp.offset = 0;
		agp.size = 0;
		agp.free = 0;
	}

	DRM_COPY_TO_USER_IOCTL(argp, agp, sizeof(agp));

	DRM_DEBUG("alloc agp, size = %d, offset = %d\n", agp.size, agp.offset);

	return retval;
}

int sis_ioctl_agp_free( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_sis_private_t *dev_priv = dev->dev_private;
	drm_sis_mem_t agp;

	if (dev_priv == NULL || dev_priv->AGPHeap == NULL)
		return DRM_ERR(EINVAL);

	DRM_COPY_FROM_USER_IOCTL(agp, (drm_sis_mem_t __user *)data, sizeof(agp));

	if (!mmBlockInHeap(dev_priv->AGPHeap, (PMemBlock)agp.free))
		return DRM_ERR(EINVAL);

	mmFreeMem((PMemBlock)agp.free);
	if (!del_alloc_set(agp.context, AGP_TYPE, agp.free))
		return DRM_ERR(EINVAL);

	DRM_DEBUG("free agp, free = 0x%lx\n", agp.free);

	return 0;
}

int sis_init_context(struct drm_device *dev, int context)
{
	int i;

	for (i = 0; i < MAX_CONTEXT ; i++) {
		if (global_ppriv[i].used &&
		    (global_ppriv[i].context == context))
			break;
	}

	if (i >= MAX_CONTEXT) {
		for (i = 0; i < MAX_CONTEXT ; i++) {
			if (!global_ppriv[i].used) {
				global_ppriv[i].context = context;
				global_ppriv[i].used = 1;
				global_ppriv[i].sets[0] = setInit();
				global_ppriv[i].sets[1] = setInit();
				DRM_DEBUG("init allocation set, socket=%d, "
				    "context = %d\n", i, context);
				break;
			}
		}
		if ((i >= MAX_CONTEXT) || (global_ppriv[i].sets[0] == NULL) ||
		    (global_ppriv[i].sets[1] == NULL))
		{
			return 0;
		}
	}
	
	return 1;
}

int sis_final_context(struct drm_device *dev, int context)
{
	int i;

	for (i=0; i<MAX_CONTEXT; i++) {
		if (global_ppriv[i].used &&
		    (global_ppriv[i].context == context))
			break;
	}

	if (i < MAX_CONTEXT) {
		set_t *set;
		ITEM_TYPE item;
		int retval;

		DRM_DEBUG("find socket %d, context = %d\n", i, context);

		/* Video Memory */
		set = global_ppriv[i].sets[0];
		retval = setFirst(set, &item);
		while (retval) {
			DRM_DEBUG("free video memory 0x%lx\n", item);
#if defined(__linux__) && defined(CONFIG_FB_SIS)
			sis_free(item);
#else
			mmFreeMem((PMemBlock)item);
#endif
			retval = setNext(set, &item);
		}
		setDestroy(set);

		/* AGP Memory */
		set = global_ppriv[i].sets[1];
		retval = setFirst(set, &item);
		while (retval) {
			DRM_DEBUG("free agp memory 0x%lx\n", item);
			mmFreeMem((PMemBlock)item);
			retval = setNext(set, &item);
		}
		setDestroy(set);

		global_ppriv[i].used = 0;	  
        }
	
	return 1;
}

void DRM(driver_register_fns)(drm_device_t *dev)
{
	dev->driver_features = DRIVER_USE_AGP | DRIVER_USE_MTRR;
	dev->fn_tbl.context_ctor = sis_init_context;
	dev->fn_tbl.context_dtor = sis_final_context;
}
>1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622
/* radeon_state.c -- State support for Radeon -*- linux-c -*-
 *
 * Copyright 2000 VA Linux Systems, Inc., Fremont, California.
 * All Rights Reserved.
 *
 * 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:
 *    Gareth Hughes <gareth@valinux.com>
 *    Kevin E. Martin <martin@valinux.com>
 */

#include "radeon.h"
#include "drmP.h"
#include "drm.h"
#include "drm_sarea.h"
#include "radeon_drm.h"
#include "radeon_drv.h"


/* ================================================================
 * Helper functions for client state checking and fixup
 */

static __inline__ int radeon_check_and_fixup_offset( drm_radeon_private_t *dev_priv,
						     drm_file_t *filp_priv,
						     u32 *offset ) {
	u32 off = *offset;
	struct drm_radeon_driver_file_fields *radeon_priv;

	if ( off >= dev_priv->fb_location &&
	     off < ( dev_priv->gart_vm_start + dev_priv->gart_size ) )
		return 0;

	radeon_priv = filp_priv->driver_priv;

	off += radeon_priv->radeon_fb_delta;

	DRM_DEBUG( "offset fixed up to 0x%x\n", off );

	if ( off < dev_priv->fb_location ||
	     off >= ( dev_priv->gart_vm_start + dev_priv->gart_size ) )
		return DRM_ERR( EINVAL );

	*offset = off;

	return 0;
}

static __inline__ int radeon_check_and_fixup_offset_user( drm_radeon_private_t *dev_priv,
							  drm_file_t *filp_priv,
							  u32 __user *offset ) {
	u32 off;

	DRM_GET_USER_UNCHECKED( off, offset );

	if ( radeon_check_and_fixup_offset( dev_priv, filp_priv, &off ) )
		return DRM_ERR( EINVAL );

	DRM_PUT_USER_UNCHECKED( offset, off );

	return 0;
}

static __inline__ int radeon_check_and_fixup_packets( drm_radeon_private_t *dev_priv,
						      drm_file_t *filp_priv,
						      int id,
						      u32 __user *data ) {
	switch ( id ) {

	case RADEON_EMIT_PP_MISC:
		if ( radeon_check_and_fixup_offset_user( dev_priv, filp_priv,
							 &data[( RADEON_RB3D_DEPTHOFFSET
								 - RADEON_PP_MISC ) / 4] ) ) {
			DRM_ERROR( "Invalid depth buffer offset\n" );
			return DRM_ERR( EINVAL );
		}
		break;

	case RADEON_EMIT_PP_CNTL:
		if ( radeon_check_and_fixup_offset_user( dev_priv, filp_priv,
							 &data[( RADEON_RB3D_COLOROFFSET
								 - RADEON_PP_CNTL ) / 4] ) ) {
			DRM_ERROR( "Invalid colour buffer offset\n" );
			return DRM_ERR( EINVAL );
		}
		break;

	case R200_EMIT_PP_TXOFFSET_0:
	case R200_EMIT_PP_TXOFFSET_1:
	case R200_EMIT_PP_TXOFFSET_2:
	case R200_EMIT_PP_TXOFFSET_3:
	case R200_EMIT_PP_TXOFFSET_4:
	case R200_EMIT_PP_TXOFFSET_5:
		if ( radeon_check_and_fixup_offset_user( dev_priv, filp_priv,
							 &data[0] ) ) {
			DRM_ERROR( "Invalid R200 texture offset\n" );
			return DRM_ERR( EINVAL );
		}
		break;

	case RADEON_EMIT_PP_TXFILTER_0:
	case RADEON_EMIT_PP_TXFILTER_1:
	case RADEON_EMIT_PP_TXFILTER_2:
		if ( radeon_check_and_fixup_offset_user( dev_priv, filp_priv,
							 &data[( RADEON_PP_TXOFFSET_0
								 - RADEON_PP_TXFILTER_0 ) / 4] ) ) {
			DRM_ERROR( "Invalid R100 texture offset\n" );
			return DRM_ERR( EINVAL );
		}
		break;

	case R200_EMIT_PP_CUBIC_OFFSETS_0:
	case R200_EMIT_PP_CUBIC_OFFSETS_1:
	case R200_EMIT_PP_CUBIC_OFFSETS_2:
	case R200_EMIT_PP_CUBIC_OFFSETS_3:
	case R200_EMIT_PP_CUBIC_OFFSETS_4:
	case R200_EMIT_PP_CUBIC_OFFSETS_5: {
		int i;
		for ( i = 0; i < 5; i++ ) {
			if ( radeon_check_and_fixup_offset_user( dev_priv,
								 filp_priv,
								 &data[i] ) ) {
				DRM_ERROR( "Invalid R200 cubic texture offset\n" );
				return DRM_ERR( EINVAL );
			}
		}
		break;
	}

	case RADEON_EMIT_RB3D_COLORPITCH:
	case RADEON_EMIT_RE_LINE_PATTERN:
	case RADEON_EMIT_SE_LINE_WIDTH:
	case RADEON_EMIT_PP_LUM_MATRIX:
	case RADEON_EMIT_PP_ROT_MATRIX_0:
	case RADEON_EMIT_RB3D_STENCILREFMASK:
	case RADEON_EMIT_SE_VPORT_XSCALE:
	case RADEON_EMIT_SE_CNTL:
	case RADEON_EMIT_SE_CNTL_STATUS:
	case RADEON_EMIT_RE_MISC:
	case RADEON_EMIT_PP_BORDER_COLOR_0:
	case RADEON_EMIT_PP_BORDER_COLOR_1:
	case RADEON_EMIT_PP_BORDER_COLOR_2:
	case RADEON_EMIT_SE_ZBIAS_FACTOR:
	case RADEON_EMIT_SE_TCL_OUTPUT_VTX_FMT:
	case RADEON_EMIT_SE_TCL_MATERIAL_EMMISSIVE_RED:
	case R200_EMIT_PP_TXCBLEND_0:
	case R200_EMIT_PP_TXCBLEND_1:
	case R200_EMIT_PP_TXCBLEND_2:
	case R200_EMIT_PP_TXCBLEND_3:
	case R200_EMIT_PP_TXCBLEND_4:
	case R200_EMIT_PP_TXCBLEND_5:
	case R200_EMIT_PP_TXCBLEND_6:
	case R200_EMIT_PP_TXCBLEND_7:
	case R200_EMIT_TCL_LIGHT_MODEL_CTL_0:
	case R200_EMIT_TFACTOR_0:
	case R200_EMIT_VTX_FMT_0:
	case R200_EMIT_VAP_CTL:
	case R200_EMIT_MATRIX_SELECT_0:
	case R200_EMIT_TEX_PROC_CTL_2:
	case R200_EMIT_TCL_UCP_VERT_BLEND_CTL:
	case R200_EMIT_PP_TXFILTER_0:
	case R200_EMIT_PP_TXFILTER_1:
	case R200_EMIT_PP_TXFILTER_2:
	case R200_EMIT_PP_TXFILTER_3:
	case R200_EMIT_PP_TXFILTER_4:
	case R200_EMIT_PP_TXFILTER_5:
	case R200_EMIT_VTE_CNTL:
	case R200_EMIT_OUTPUT_VTX_COMP_SEL:
	case R200_EMIT_PP_TAM_DEBUG3:
	case R200_EMIT_PP_CNTL_X:
	case R200_EMIT_RB3D_DEPTHXY_OFFSET:
	case R200_EMIT_RE_AUX_SCISSOR_CNTL:
	case R200_EMIT_RE_SCISSOR_TL_0:
	case R200_EMIT_RE_SCISSOR_TL_1:
	case R200_EMIT_RE_SCISSOR_TL_2:
	case R200_EMIT_SE_VAP_CNTL_STATUS:
	case R200_EMIT_SE_VTX_STATE_CNTL:
	case R200_EMIT_RE_POINTSIZE:
	case R200_EMIT_TCL_INPUT_VTX_VECTOR_ADDR_0:
	case R200_EMIT_PP_CUBIC_FACES_0:
	case R200_EMIT_PP_CUBIC_FACES_1:
	case R200_EMIT_PP_CUBIC_FACES_2:
	case R200_EMIT_PP_CUBIC_FACES_3:
	case R200_EMIT_PP_CUBIC_FACES_4:
	case R200_EMIT_PP_CUBIC_FACES_5:
	case RADEON_EMIT_PP_TEX_SIZE_0:
	case RADEON_EMIT_PP_TEX_SIZE_1:
	case RADEON_EMIT_PP_TEX_SIZE_2:
	case R200_EMIT_RB3D_BLENDCOLOR:
		/* These packets don't contain memory offsets */
		break;

	default:
		DRM_ERROR( "Unknown state packet ID %d\n", id );
		return DRM_ERR( EINVAL );
	}

	return 0;
}

static __inline__ int radeon_check_and_fixup_packet3( drm_radeon_private_t *dev_priv,
						      drm_file_t *filp_priv,
						      drm_radeon_cmd_buffer_t *cmdbuf,
						      unsigned int *cmdsz ) {
	u32 tmp[4];
	u32 __user *cmd = (u32 __user *)cmdbuf->buf;

	if ( DRM_COPY_FROM_USER_UNCHECKED( tmp, cmd, sizeof( tmp ) ) ) {
		DRM_ERROR( "Failed to copy data from user space\n" );
		return DRM_ERR( EFAULT );
	}

	*cmdsz = 2 + ( ( tmp[0] & RADEON_CP_PACKET_COUNT_MASK ) >> 16 );

	if ( ( tmp[0] & 0xc0000000 ) != RADEON_CP_PACKET3 ) {
		DRM_ERROR( "Not a type 3 packet\n" );
		return DRM_ERR( EINVAL );
	}

	if ( 4 * *cmdsz > cmdbuf->bufsz ) {
		DRM_ERROR( "Packet size larger than size of data provided\n" );
		return DRM_ERR( EINVAL );
	}

	/* Check client state and fix it up if necessary */
	if ( tmp[0] & 0x8000 ) { /* MSB of opcode: next DWORD GUI_CNTL */
		u32 offset;

		if ( tmp[1] & ( RADEON_GMC_SRC_PITCH_OFFSET_CNTL
			      | RADEON_GMC_DST_PITCH_OFFSET_CNTL ) ) {
			offset = tmp[2] << 10;
			if ( radeon_check_and_fixup_offset( dev_priv, filp_priv, &offset ) ) {
				DRM_ERROR( "Invalid first packet offset\n" );
				return DRM_ERR( EINVAL );
			}
			tmp[2] = ( tmp[2] & 0xffc00000 ) | offset >> 10;
		}

		if ( ( tmp[1] & RADEON_GMC_SRC_PITCH_OFFSET_CNTL ) &&
		     ( tmp[1] & RADEON_GMC_DST_PITCH_OFFSET_CNTL ) ) {
			offset = tmp[3] << 10;
			if ( radeon_check_and_fixup_offset( dev_priv, filp_priv, &offset ) ) {
				DRM_ERROR( "Invalid second packet offset\n" );
				return DRM_ERR( EINVAL );
			}
			tmp[3] = ( tmp[3] & 0xffc00000 ) | offset >> 10;
		}

		if ( DRM_COPY_TO_USER_UNCHECKED( cmd, tmp, sizeof( tmp ) ) ) {
			DRM_ERROR( "Failed to copy data to user space\n" );
			return DRM_ERR( EFAULT );
		}
	}

	return 0;
}


/* ================================================================
 * CP hardware state programming functions
 */

static __inline__ void radeon_emit_clip_rect( drm_radeon_private_t *dev_priv,
					  drm_clip_rect_t *box )
{
	RING_LOCALS;

	DRM_DEBUG( "   box:  x1=%d y1=%d  x2=%d y2=%d\n",
		   box->x1, box->y1, box->x2, box->y2 );

	BEGIN_RING( 4 );
	OUT_RING( CP_PACKET0( RADEON_RE_TOP_LEFT, 0 ) );
	OUT_RING( (box->y1 << 16) | box->x1 );
	OUT_RING( CP_PACKET0( RADEON_RE_WIDTH_HEIGHT, 0 ) );
	OUT_RING( ((box->y2 - 1) << 16) | (box->x2 - 1) );
	ADVANCE_RING();
}

/* Emit 1.1 state
 */
static int radeon_emit_state( drm_radeon_private_t *dev_priv,
			      drm_file_t *filp_priv,
			      drm_radeon_context_regs_t *ctx,
			      drm_radeon_texture_regs_t *tex,
			      unsigned int dirty )
{
	RING_LOCALS;
	DRM_DEBUG( "dirty=0x%08x\n", dirty );

	if ( dirty & RADEON_UPLOAD_CONTEXT ) {
		if ( radeon_check_and_fixup_offset( dev_priv, filp_priv,
						    &ctx->rb3d_depthoffset ) ) {
			DRM_ERROR( "Invalid depth buffer offset\n" );
			return DRM_ERR( EINVAL );
		}

		if ( radeon_check_and_fixup_offset( dev_priv, filp_priv,
						    &ctx->rb3d_coloroffset ) ) {
			DRM_ERROR( "Invalid depth buffer offset\n" );
			return DRM_ERR( EINVAL );
		}

		BEGIN_RING( 14 );
		OUT_RING( CP_PACKET0( RADEON_PP_MISC, 6 ) );
		OUT_RING( ctx->pp_misc );
		OUT_RING( ctx->pp_fog_color );
		OUT_RING( ctx->re_solid_color );
		OUT_RING( ctx->rb3d_blendcntl );
		OUT_RING( ctx->rb3d_depthoffset );
		OUT_RING( ctx->rb3d_depthpitch );
		OUT_RING( ctx->rb3d_zstencilcntl );
		OUT_RING( CP_PACKET0( RADEON_PP_CNTL, 2 ) );
		OUT_RING( ctx->pp_cntl );
		OUT_RING( ctx->rb3d_cntl );
		OUT_RING( ctx->rb3d_coloroffset );
		OUT_RING( CP_PACKET0( RADEON_RB3D_COLORPITCH, 0 ) );
		OUT_RING( ctx->rb3d_colorpitch );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_VERTFMT ) {
		BEGIN_RING( 2 );
		OUT_RING( CP_PACKET0( RADEON_SE_COORD_FMT, 0 ) );
		OUT_RING( ctx->se_coord_fmt );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_LINE ) {
		BEGIN_RING( 5 );
		OUT_RING( CP_PACKET0( RADEON_RE_LINE_PATTERN, 1 ) );
		OUT_RING( ctx->re_line_pattern );
		OUT_RING( ctx->re_line_state );
		OUT_RING( CP_PACKET0( RADEON_SE_LINE_WIDTH, 0 ) );
		OUT_RING( ctx->se_line_width );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_BUMPMAP ) {
		BEGIN_RING( 5 );
		OUT_RING( CP_PACKET0( RADEON_PP_LUM_MATRIX, 0 ) );
		OUT_RING( ctx->pp_lum_matrix );
		OUT_RING( CP_PACKET0( RADEON_PP_ROT_MATRIX_0, 1 ) );
		OUT_RING( ctx->pp_rot_matrix_0 );
		OUT_RING( ctx->pp_rot_matrix_1 );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_MASKS ) {
		BEGIN_RING( 4 );
		OUT_RING( CP_PACKET0( RADEON_RB3D_STENCILREFMASK, 2 ) );
		OUT_RING( ctx->rb3d_stencilrefmask );
		OUT_RING( ctx->rb3d_ropcntl );
		OUT_RING( ctx->rb3d_planemask );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_VIEWPORT ) {
		BEGIN_RING( 7 );
		OUT_RING( CP_PACKET0( RADEON_SE_VPORT_XSCALE, 5 ) );
		OUT_RING( ctx->se_vport_xscale );
		OUT_RING( ctx->se_vport_xoffset );
		OUT_RING( ctx->se_vport_yscale );
		OUT_RING( ctx->se_vport_yoffset );
		OUT_RING( ctx->se_vport_zscale );
		OUT_RING( ctx->se_vport_zoffset );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_SETUP ) {
		BEGIN_RING( 4 );
		OUT_RING( CP_PACKET0( RADEON_SE_CNTL, 0 ) );
		OUT_RING( ctx->se_cntl );
		OUT_RING( CP_PACKET0( RADEON_SE_CNTL_STATUS, 0 ) );
		OUT_RING( ctx->se_cntl_status );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_MISC ) {
		BEGIN_RING( 2 );
		OUT_RING( CP_PACKET0( RADEON_RE_MISC, 0 ) );
		OUT_RING( ctx->re_misc );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_TEX0 ) {
		if ( radeon_check_and_fixup_offset( dev_priv, filp_priv,
						    &tex[0].pp_txoffset ) ) {
			DRM_ERROR( "Invalid texture offset for unit 0\n" );
			return DRM_ERR( EINVAL );
		}

		BEGIN_RING( 9 );
		OUT_RING( CP_PACKET0( RADEON_PP_TXFILTER_0, 5 ) );
		OUT_RING( tex[0].pp_txfilter );
		OUT_RING( tex[0].pp_txformat );
		OUT_RING( tex[0].pp_txoffset );
		OUT_RING( tex[0].pp_txcblend );
		OUT_RING( tex[0].pp_txablend );
		OUT_RING( tex[0].pp_tfactor );
		OUT_RING( CP_PACKET0( RADEON_PP_BORDER_COLOR_0, 0 ) );
		OUT_RING( tex[0].pp_border_color );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_TEX1 ) {
		if ( radeon_check_and_fixup_offset( dev_priv, filp_priv,
						    &tex[1].pp_txoffset ) ) {
			DRM_ERROR( "Invalid texture offset for unit 1\n" );
			return DRM_ERR( EINVAL );
		}

		BEGIN_RING( 9 );
		OUT_RING( CP_PACKET0( RADEON_PP_TXFILTER_1, 5 ) );
		OUT_RING( tex[1].pp_txfilter );
		OUT_RING( tex[1].pp_txformat );
		OUT_RING( tex[1].pp_txoffset );
		OUT_RING( tex[1].pp_txcblend );
		OUT_RING( tex[1].pp_txablend );
		OUT_RING( tex[1].pp_tfactor );
		OUT_RING( CP_PACKET0( RADEON_PP_BORDER_COLOR_1, 0 ) );
		OUT_RING( tex[1].pp_border_color );
		ADVANCE_RING();
	}

	if ( dirty & RADEON_UPLOAD_TEX2 ) {
		if ( radeon_check_and_fixup_offset( dev_priv, filp_priv,
						    &tex[2].pp_txoffset ) ) {
			DRM_ERROR( "Invalid texture offset for unit 2\n" );
			return DRM_ERR( EINVAL );
		}

		BEGIN_RING( 9 );
		OUT_RING( CP_PACKET0( RADEON_PP_TXFILTER_2, 5 ) );
		OUT_RING( tex[2].pp_txfilter );
		OUT_RING( tex[2].pp_txformat );
		OUT_RING( tex[2].pp_txoffset );
		OUT_RING( tex[2].pp_txcblend );
		OUT_RING( tex[2].pp_txablend );
		OUT_RING( tex[2].pp_tfactor );
		OUT_RING( CP_PACKET0( RADEON_PP_BORDER_COLOR_2, 0 ) );
		OUT_RING( tex[2].pp_border_color );
		ADVANCE_RING();
	}

	return 0;
}

/* Emit 1.2 state
 */
static int radeon_emit_state2( drm_radeon_private_t *dev_priv,
			       drm_file_t *filp_priv,
			       drm_radeon_state_t *state )
{
	RING_LOCALS;

	if (state->dirty & RADEON_UPLOAD_ZBIAS) {
		BEGIN_RING( 3 );
		OUT_RING( CP_PACKET0( RADEON_SE_ZBIAS_FACTOR, 1 ) );
		OUT_RING( state->context2.se_zbias_factor ); 
		OUT_RING( state->context2.se_zbias_constant ); 
		ADVANCE_RING();
	}

	return radeon_emit_state( dev_priv, filp_priv, &state->context,
			   state->tex, state->dirty );
}

/* New (1.3) state mechanism.  3 commands (packet, scalar, vector) in
 * 1.3 cmdbuffers allow all previous state to be updated as well as
 * the tcl scalar and vector areas.  
 */
static struct { 
	int start; 
	int len; 
	const char *name;
} packet[RADEON_MAX_STATE_PACKETS] = {
	{ RADEON_PP_MISC,7,"RADEON_PP_MISC" },
	{ RADEON_PP_CNTL,3,"RADEON_PP_CNTL" },
	{ RADEON_RB3D_COLORPITCH,1,"RADEON_RB3D_COLORPITCH" },
	{ RADEON_RE_LINE_PATTERN,2,"RADEON_RE_LINE_PATTERN" },
	{ RADEON_SE_LINE_WIDTH,1,"RADEON_SE_LINE_WIDTH" },
	{ RADEON_PP_LUM_MATRIX,1,"RADEON_PP_LUM_MATRIX" },
	{ RADEON_PP_ROT_MATRIX_0,2,"RADEON_PP_ROT_MATRIX_0" },
	{ RADEON_RB3D_STENCILREFMASK,3,"RADEON_RB3D_STENCILREFMASK" },
	{ RADEON_SE_VPORT_XSCALE,6,"RADEON_SE_VPORT_XSCALE" },
	{ RADEON_SE_CNTL,2,"RADEON_SE_CNTL" },
	{ RADEON_SE_CNTL_STATUS,1,"RADEON_SE_CNTL_STATUS" },
	{ RADEON_RE_MISC,1,"RADEON_RE_MISC" },
	{ RADEON_PP_TXFILTER_0,6,"RADEON_PP_TXFILTER_0" },
	{ RADEON_PP_BORDER_COLOR_0,1,"RADEON_PP_BORDER_COLOR_0" },
	{ RADEON_PP_TXFILTER_1,6,"RADEON_PP_TXFILTER_1" },
	{ RADEON_PP_BORDER_COLOR_1,1,"RADEON_PP_BORDER_COLOR_1" },
	{ RADEON_PP_TXFILTER_2,6,"RADEON_PP_TXFILTER_2" },
	{ RADEON_PP_BORDER_COLOR_2,1,"RADEON_PP_BORDER_COLOR_2" },
	{ RADEON_SE_ZBIAS_FACTOR,2,"RADEON_SE_ZBIAS_FACTOR" },
	{ RADEON_SE_TCL_OUTPUT_VTX_FMT,11,"RADEON_SE_TCL_OUTPUT_VTX_FMT" },
	{ RADEON_SE_TCL_MATERIAL_EMMISSIVE_RED,17,"RADEON_SE_TCL_MATERIAL_EMMISSIVE_RED" },
	{ R200_PP_TXCBLEND_0, 4, "R200_PP_TXCBLEND_0" },
	{ R200_PP_TXCBLEND_1, 4, "R200_PP_TXCBLEND_1" },
	{ R200_PP_TXCBLEND_2, 4, "R200_PP_TXCBLEND_2" },
	{ R200_PP_TXCBLEND_3, 4, "R200_PP_TXCBLEND_3" },
	{ R200_PP_TXCBLEND_4, 4, "R200_PP_TXCBLEND_4" },
	{ R200_PP_TXCBLEND_5, 4, "R200_PP_TXCBLEND_5" },
	{ R200_PP_TXCBLEND_6, 4, "R200_PP_TXCBLEND_6" },
	{ R200_PP_TXCBLEND_7, 4, "R200_PP_TXCBLEND_7" },
	{ R200_SE_TCL_LIGHT_MODEL_CTL_0, 6, "R200_SE_TCL_LIGHT_MODEL_CTL_0" },
	{ R200_PP_TFACTOR_0, 6, "R200_PP_TFACTOR_0" },
	{ R200_SE_VTX_FMT_0, 4, "R200_SE_VTX_FMT_0" },
	{ R200_SE_VAP_CNTL, 1, "R200_SE_VAP_CNTL" },
	{ R200_SE_TCL_MATRIX_SEL_0, 5, "R200_SE_TCL_MATRIX_SEL_0" },
	{ R200_SE_TCL_TEX_PROC_CTL_2, 5, "R200_SE_TCL_TEX_PROC_CTL_2" },
	{ R200_SE_TCL_UCP_VERT_BLEND_CTL, 1, "R200_SE_TCL_UCP_VERT_BLEND_CTL" },
	{ R200_PP_TXFILTER_0, 6, "R200_PP_TXFILTER_0" },
	{ R200_PP_TXFILTER_1, 6, "R200_PP_TXFILTER_1" },
	{ R200_PP_TXFILTER_2, 6, "R200_PP_TXFILTER_2" },
	{ R200_PP_TXFILTER_3, 6, "R200_PP_TXFILTER_3" },
	{ R200_PP_TXFILTER_4, 6, "R200_PP_TXFILTER_4" },
	{ R200_PP_TXFILTER_5, 6, "R200_PP_TXFILTER_5" },
	{ R200_PP_TXOFFSET_0, 1, "R200_PP_TXOFFSET_0" },
	{ R200_PP_TXOFFSET_1, 1, "R200_PP_TXOFFSET_1" },
	{ R200_PP_TXOFFSET_2, 1, "R200_PP_TXOFFSET_2" },
	{ R200_PP_TXOFFSET_3, 1, "R200_PP_TXOFFSET_3" },
	{ R200_PP_TXOFFSET_4, 1, "R200_PP_TXOFFSET_4" },
	{ R200_PP_TXOFFSET_5, 1, "R200_PP_TXOFFSET_5" },
	{ R200_SE_VTE_CNTL, 1, "R200_SE_VTE_CNTL" },
	{ R200_SE_TCL_OUTPUT_VTX_COMP_SEL, 1, "R200_SE_TCL_OUTPUT_VTX_COMP_SEL" },
	{ R200_PP_TAM_DEBUG3, 1, "R200_PP_TAM_DEBUG3" },
	{ R200_PP_CNTL_X, 1, "R200_PP_CNTL_X" }, 
	{ R200_RB3D_DEPTHXY_OFFSET, 1, "R200_RB3D_DEPTHXY_OFFSET" }, 
	{ R200_RE_AUX_SCISSOR_CNTL, 1, "R200_RE_AUX_SCISSOR_CNTL" }, 
	{ R200_RE_SCISSOR_TL_0, 2, "R200_RE_SCISSOR_TL_0" }, 
	{ R200_RE_SCISSOR_TL_1, 2, "R200_RE_SCISSOR_TL_1" }, 
	{ R200_RE_SCISSOR_TL_2, 2, "R200_RE_SCISSOR_TL_2" }, 
	{ R200_SE_VAP_CNTL_STATUS, 1, "R200_SE_VAP_CNTL_STATUS" }, 
	{ R200_SE_VTX_STATE_CNTL, 1, "R200_SE_VTX_STATE_CNTL" }, 
	{ R200_RE_POINTSIZE, 1, "R200_RE_POINTSIZE" }, 
	{ R200_SE_TCL_INPUT_VTX_VECTOR_ADDR_0, 4, "R200_SE_TCL_INPUT_VTX_VECTOR_ADDR_0" },
	{ R200_PP_CUBIC_FACES_0, 1, "R200_PP_CUBIC_FACES_0" }, /* 61 */
	{ R200_PP_CUBIC_OFFSET_F1_0, 5, "R200_PP_CUBIC_OFFSET_F1_0" }, /* 62 */
	{ R200_PP_CUBIC_FACES_1, 1, "R200_PP_CUBIC_FACES_1" },
	{ R200_PP_CUBIC_OFFSET_F1_1, 5, "R200_PP_CUBIC_OFFSET_F1_1" },
	{ R200_PP_CUBIC_FACES_2, 1, "R200_PP_CUBIC_FACES_2" },
	{ R200_PP_CUBIC_OFFSET_F1_2, 5, "R200_PP_CUBIC_OFFSET_F1_2" },
	{ R200_PP_CUBIC_FACES_3, 1, "R200_PP_CUBIC_FACES_3" },
	{ R200_PP_CUBIC_OFFSET_F1_3, 5, "R200_PP_CUBIC_OFFSET_F1_3" },
	{ R200_PP_CUBIC_FACES_4, 1, "R200_PP_CUBIC_FACES_4" },
	{ R200_PP_CUBIC_OFFSET_F1_4, 5, "R200_PP_CUBIC_OFFSET_F1_4" },
	{ R200_PP_CUBIC_FACES_5, 1, "R200_PP_CUBIC_FACES_5" },
	{ R200_PP_CUBIC_OFFSET_F1_5, 5, "R200_PP_CUBIC_OFFSET_F1_5" },
	{ RADEON_PP_TEX_SIZE_0, 2, "RADEON_PP_TEX_SIZE_0" },
	{ RADEON_PP_TEX_SIZE_1, 2, "RADEON_PP_TEX_SIZE_1" },
	{ RADEON_PP_TEX_SIZE_2, 2, "RADEON_PP_TEX_SIZE_2" },
	{ R200_RB3D_BLENDCOLOR, 3, "R200_RB3D_BLENDCOLOR" },
};



/* ================================================================
 * Performance monitoring functions
 */

static void radeon_clear_box( drm_radeon_private_t *dev_priv,
			      int x, int y, int w, int h,
			      int r, int g, int b )
{
	u32 color;
	RING_LOCALS;

	x += dev_priv->sarea_priv->boxes[0].x1;
	y += dev_priv->sarea_priv->boxes[0].y1;

	switch ( dev_priv->color_fmt ) {
	case RADEON_COLOR_FORMAT_RGB565:
		color = (((r & 0xf8) << 8) |
			 ((g & 0xfc) << 3) |
			 ((b & 0xf8) >> 3));
		break;
	case RADEON_COLOR_FORMAT_ARGB8888:
	default:
		color = (((0xff) << 24) | (r << 16) | (g <<  8) | b);
		break;
	}

	BEGIN_RING( 4 );
	RADEON_WAIT_UNTIL_3D_IDLE();		
	OUT_RING( CP_PACKET0( RADEON_DP_WRITE_MASK, 0 ) );
	OUT_RING( 0xffffffff );
	ADVANCE_RING();

	BEGIN_RING( 6 );

	OUT_RING( CP_PACKET3( RADEON_CNTL_PAINT_MULTI, 4 ) );
	OUT_RING( RADEON_GMC_DST_PITCH_OFFSET_CNTL |
		  RADEON_GMC_BRUSH_SOLID_COLOR |
		  (dev_priv->color_fmt << 8) |
		  RADEON_GMC_SRC_DATATYPE_COLOR |
		  RADEON_ROP3_P |
		  RADEON_GMC_CLR_CMP_CNTL_DIS );

 	if ( dev_priv->page_flipping && dev_priv->current_page == 1 ) { 
		OUT_RING( dev_priv->front_pitch_offset );
 	} else {	 
		OUT_RING( dev_priv->back_pitch_offset );
 	} 

	OUT_RING( color );

	OUT_RING( (x << 16) | y );
	OUT_RING( (w << 16) | h );

	ADVANCE_RING();
}

static void radeon_cp_performance_boxes( drm_radeon_private_t *dev_priv )
{
	/* Collapse various things into a wait flag -- trying to
	 * guess if userspase slept -- better just to have them tell us.
	 */
	if (dev_priv->stats.last_frame_reads > 1 ||
	    dev_priv->stats.last_clear_reads > dev_priv->stats.clears) {
		dev_priv->stats.boxes |= RADEON_BOX_WAIT_IDLE;
	}

	if (dev_priv->stats.freelist_loops) {
		dev_priv->stats.boxes |= RADEON_BOX_WAIT_IDLE;
	}

	/* Purple box for page flipping
	 */
	if ( dev_priv->stats.boxes & RADEON_BOX_FLIP ) 
		radeon_clear_box( dev_priv, 4, 4, 8, 8, 255, 0, 255 );

	/* Red box if we have to wait for idle at any point
	 */
	if ( dev_priv->stats.boxes & RADEON_BOX_WAIT_IDLE ) 
		radeon_clear_box( dev_priv, 16, 4, 8, 8, 255, 0, 0 );

	/* Blue box: lost context?
	 */

	/* Yellow box for texture swaps
	 */
	if ( dev_priv->stats.boxes & RADEON_BOX_TEXTURE_LOAD ) 
		radeon_clear_box( dev_priv, 40, 4, 8, 8, 255, 255, 0 );

	/* Green box if hardware never idles (as far as we can tell)
	 */
	if ( !(dev_priv->stats.boxes & RADEON_BOX_DMA_IDLE) ) 
		radeon_clear_box( dev_priv, 64, 4, 8, 8, 0, 255, 0 );


	/* Draw bars indicating number of buffers allocated 
	 * (not a great measure, easily confused)
	 */
	if (dev_priv->stats.requested_bufs) {
		if (dev_priv->stats.requested_bufs > 100)
			dev_priv->stats.requested_bufs = 100;

		radeon_clear_box( dev_priv, 4, 16,  
				  dev_priv->stats.requested_bufs, 4,
				  196, 128, 128 );
	}

	memset( &dev_priv->stats, 0, sizeof(dev_priv->stats) );

}
/* ================================================================
 * CP command dispatch functions
 */

static void radeon_cp_dispatch_clear( drm_device_t *dev,
				      drm_radeon_clear_t *clear,
				      drm_radeon_clear_rect_t *depth_boxes )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_radeon_sarea_t *sarea_priv = dev_priv->sarea_priv;
	drm_radeon_depth_clear_t *depth_clear = &dev_priv->depth_clear;
	int nbox = sarea_priv->nbox;
	drm_clip_rect_t *pbox = sarea_priv->boxes;
	unsigned int flags = clear->flags;
	u32 rb3d_cntl = 0, rb3d_stencilrefmask= 0;
	int i;
	RING_LOCALS;
	DRM_DEBUG( "flags = 0x%x\n", flags );

	dev_priv->stats.clears++;

	if ( dev_priv->page_flipping && dev_priv->current_page == 1 ) {
		unsigned int tmp = flags;

		flags &= ~(RADEON_FRONT | RADEON_BACK);
		if ( tmp & RADEON_FRONT ) flags |= RADEON_BACK;
		if ( tmp & RADEON_BACK )  flags |= RADEON_FRONT;
	}

	if ( flags & (RADEON_FRONT | RADEON_BACK) ) {

		BEGIN_RING( 4 );

		/* Ensure the 3D stream is idle before doing a
		 * 2D fill to clear the front or back buffer.
		 */
		RADEON_WAIT_UNTIL_3D_IDLE();
		
		OUT_RING( CP_PACKET0( RADEON_DP_WRITE_MASK, 0 ) );
		OUT_RING( clear->color_mask );

		ADVANCE_RING();

		/* Make sure we restore the 3D state next time.
		 */
		dev_priv->sarea_priv->ctx_owner = 0;

		for ( i = 0 ; i < nbox ; i++ ) {
			int x = pbox[i].x1;
			int y = pbox[i].y1;
			int w = pbox[i].x2 - x;
			int h = pbox[i].y2 - y;

			DRM_DEBUG( "dispatch clear %d,%d-%d,%d flags 0x%x\n",
				   x, y, w, h, flags );

			if ( flags & RADEON_FRONT ) {
				BEGIN_RING( 6 );
				
				OUT_RING( CP_PACKET3( RADEON_CNTL_PAINT_MULTI, 4 ) );
				OUT_RING( RADEON_GMC_DST_PITCH_OFFSET_CNTL |
					  RADEON_GMC_BRUSH_SOLID_COLOR |
					  (dev_priv->color_fmt << 8) |
					  RADEON_GMC_SRC_DATATYPE_COLOR |
					  RADEON_ROP3_P |
					  RADEON_GMC_CLR_CMP_CNTL_DIS );

				OUT_RING( dev_priv->front_pitch_offset );
				OUT_RING( clear->clear_color );
				
				OUT_RING( (x << 16) | y );
				OUT_RING( (w << 16) | h );
				
				ADVANCE_RING();
			}
			
			if ( flags & RADEON_BACK ) {
				BEGIN_RING( 6 );
				
				OUT_RING( CP_PACKET3( RADEON_CNTL_PAINT_MULTI, 4 ) );
				OUT_RING( RADEON_GMC_DST_PITCH_OFFSET_CNTL |
					  RADEON_GMC_BRUSH_SOLID_COLOR |
					  (dev_priv->color_fmt << 8) |
					  RADEON_GMC_SRC_DATATYPE_COLOR |
					  RADEON_ROP3_P |
					  RADEON_GMC_CLR_CMP_CNTL_DIS );
				
				OUT_RING( dev_priv->back_pitch_offset );
				OUT_RING( clear->clear_color );

				OUT_RING( (x << 16) | y );
				OUT_RING( (w << 16) | h );

				ADVANCE_RING();
			}
		}
	}

	/* We have to clear the depth and/or stencil buffers by
	 * rendering a quad into just those buffers.  Thus, we have to
	 * make sure the 3D engine is configured correctly.
	 */
	if ( dev_priv->is_r200 &&
	     (flags & (RADEON_DEPTH | RADEON_STENCIL)) ) {

		int tempPP_CNTL;
		int tempRE_CNTL;
		int tempRB3D_CNTL;
		int tempRB3D_ZSTENCILCNTL;
		int tempRB3D_STENCILREFMASK;
		int tempRB3D_PLANEMASK;
		int tempSE_CNTL;
		int tempSE_VTE_CNTL;
		int tempSE_VTX_FMT_0;
		int tempSE_VTX_FMT_1;
		int tempSE_VAP_CNTL;
		int tempRE_AUX_SCISSOR_CNTL;

		tempPP_CNTL = 0;
		tempRE_CNTL = 0;

		tempRB3D_CNTL = depth_clear->rb3d_cntl;
		tempRB3D_CNTL &= ~(1<<15); /* unset radeon magic flag */

		tempRB3D_ZSTENCILCNTL = depth_clear->rb3d_zstencilcntl;
		tempRB3D_STENCILREFMASK = 0x0;

		tempSE_CNTL = depth_clear->se_cntl;



		/* Disable TCL */

		tempSE_VAP_CNTL = (/* SE_VAP_CNTL__FORCE_W_TO_ONE_MASK |  */
				   (0x9 << SE_VAP_CNTL__VF_MAX_VTX_NUM__SHIFT));

		tempRB3D_PLANEMASK = 0x0;

		tempRE_AUX_SCISSOR_CNTL = 0x0;

		tempSE_VTE_CNTL =
			SE_VTE_CNTL__VTX_XY_FMT_MASK |
			SE_VTE_CNTL__VTX_Z_FMT_MASK;

		/* Vertex format (X, Y, Z, W)*/
		tempSE_VTX_FMT_0 =
			SE_VTX_FMT_0__VTX_Z0_PRESENT_MASK |
			SE_VTX_FMT_0__VTX_W0_PRESENT_MASK;
		tempSE_VTX_FMT_1 = 0x0;


		/* 
		 * Depth buffer specific enables 
		 */
		if (flags & RADEON_DEPTH) {
			/* Enable depth buffer */
			tempRB3D_CNTL |= RADEON_Z_ENABLE;
		} else {
			/* Disable depth buffer */
			tempRB3D_CNTL &= ~RADEON_Z_ENABLE;
		}

		/* 
		 * Stencil buffer specific enables
		 */
		if ( flags & RADEON_STENCIL ) {
			tempRB3D_CNTL |=  RADEON_STENCIL_ENABLE;
			tempRB3D_STENCILREFMASK = clear->depth_mask; 
		} else {
			tempRB3D_CNTL &= ~RADEON_STENCIL_ENABLE;
			tempRB3D_STENCILREFMASK = 0x00000000;
		}

		BEGIN_RING( 26 );
		RADEON_WAIT_UNTIL_2D_IDLE();

		OUT_RING_REG( RADEON_PP_CNTL, tempPP_CNTL );
		OUT_RING_REG( R200_RE_CNTL, tempRE_CNTL );
		OUT_RING_REG( RADEON_RB3D_CNTL, tempRB3D_CNTL );
		OUT_RING_REG( RADEON_RB3D_ZSTENCILCNTL,
			      tempRB3D_ZSTENCILCNTL );
		OUT_RING_REG( RADEON_RB3D_STENCILREFMASK, 
			      tempRB3D_STENCILREFMASK );
		OUT_RING_REG( RADEON_RB3D_PLANEMASK, tempRB3D_PLANEMASK );
		OUT_RING_REG( RADEON_SE_CNTL, tempSE_CNTL );
		OUT_RING_REG( R200_SE_VTE_CNTL, tempSE_VTE_CNTL );
		OUT_RING_REG( R200_SE_VTX_FMT_0, tempSE_VTX_FMT_0 );
		OUT_RING_REG( R200_SE_VTX_FMT_1, tempSE_VTX_FMT_1 );
		OUT_RING_REG( R200_SE_VAP_CNTL, tempSE_VAP_CNTL );
		OUT_RING_REG( R200_RE_AUX_SCISSOR_CNTL, 
			      tempRE_AUX_SCISSOR_CNTL );
		ADVANCE_RING();

		/* Make sure we restore the 3D state next time.
		 */
		dev_priv->sarea_priv->ctx_owner = 0;

		for ( i = 0 ; i < nbox ; i++ ) {
			
			/* Funny that this should be required -- 
			 *  sets top-left?
			 */
			radeon_emit_clip_rect( dev_priv,
					       &sarea_priv->boxes[i] );

			BEGIN_RING( 14 );
			OUT_RING( CP_PACKET3( R200_3D_DRAW_IMMD_2, 12 ) );
			OUT_RING( (RADEON_PRIM_TYPE_RECT_LIST |
				   RADEON_PRIM_WALK_RING |
				   (3 << RADEON_NUM_VERTICES_SHIFT)) );
			OUT_RING( depth_boxes[i].ui[CLEAR_X1] );
			OUT_RING( depth_boxes[i].ui[CLEAR_Y1] );
			OUT_RING( depth_boxes[i].ui[CLEAR_DEPTH] );
			OUT_RING( 0x3f800000 );
			OUT_RING( depth_boxes[i].ui[CLEAR_X1] );
			OUT_RING( depth_boxes[i].ui[CLEAR_Y2] );
			OUT_RING( depth_boxes[i].ui[CLEAR_DEPTH] );
			OUT_RING( 0x3f800000 );
			OUT_RING( depth_boxes[i].ui[CLEAR_X2] );
			OUT_RING( depth_boxes[i].ui[CLEAR_Y2] );
			OUT_RING( depth_boxes[i].ui[CLEAR_DEPTH] );
			OUT_RING( 0x3f800000 );
			ADVANCE_RING();
		}
	} 
	else if ( (flags & (RADEON_DEPTH | RADEON_STENCIL)) ) {

		rb3d_cntl = depth_clear->rb3d_cntl;

		if ( flags & RADEON_DEPTH ) {
			rb3d_cntl |=  RADEON_Z_ENABLE;
		} else {
			rb3d_cntl &= ~RADEON_Z_ENABLE;
		}

		if ( flags & RADEON_STENCIL ) {
			rb3d_cntl |=  RADEON_STENCIL_ENABLE;
			rb3d_stencilrefmask = clear->depth_mask; /* misnamed field */
		} else {
			rb3d_cntl &= ~RADEON_STENCIL_ENABLE;
			rb3d_stencilrefmask = 0x00000000;
		}

		BEGIN_RING( 13 );
		RADEON_WAIT_UNTIL_2D_IDLE();

		OUT_RING( CP_PACKET0( RADEON_PP_CNTL, 1 ) );
		OUT_RING( 0x00000000 );
		OUT_RING( rb3d_cntl );
		
		OUT_RING_REG( RADEON_RB3D_ZSTENCILCNTL,
			      depth_clear->rb3d_zstencilcntl );
		OUT_RING_REG( RADEON_RB3D_STENCILREFMASK,
			      rb3d_stencilrefmask );
		OUT_RING_REG( RADEON_RB3D_PLANEMASK,
			      0x00000000 );
		OUT_RING_REG( RADEON_SE_CNTL,
			      depth_clear->se_cntl );
		ADVANCE_RING();

		/* Make sure we restore the 3D state next time.
		 */
		dev_priv->sarea_priv->ctx_owner = 0;

		for ( i = 0 ; i < nbox ; i++ ) {
			
			/* Funny that this should be required -- 
			 *  sets top-left?
			 */
			radeon_emit_clip_rect( dev_priv,
					       &sarea_priv->boxes[i] );

			BEGIN_RING( 15 );

			OUT_RING( CP_PACKET3( RADEON_3D_DRAW_IMMD, 13 ) );
			OUT_RING( RADEON_VTX_Z_PRESENT |
				  RADEON_VTX_PKCOLOR_PRESENT);
			OUT_RING( (RADEON_PRIM_TYPE_RECT_LIST |
				   RADEON_PRIM_WALK_RING |
				   RADEON_MAOS_ENABLE |
				   RADEON_VTX_FMT_RADEON_MODE |
				   (3 << RADEON_NUM_VERTICES_SHIFT)) );


			OUT_RING( depth_boxes[i].ui[CLEAR_X1] );
			OUT_RING( depth_boxes[i].ui[CLEAR_Y1] );
			OUT_RING( depth_boxes[i].ui[CLEAR_DEPTH] );
			OUT_RING( 0x0 );

			OUT_RING( depth_boxes[i].ui[CLEAR_X1] );
			OUT_RING( depth_boxes[i].ui[CLEAR_Y2] );
			OUT_RING( depth_boxes[i].ui[CLEAR_DEPTH] );
			OUT_RING( 0x0 );

			OUT_RING( depth_boxes[i].ui[CLEAR_X2] );
			OUT_RING( depth_boxes[i].ui[CLEAR_Y2] );
			OUT_RING( depth_boxes[i].ui[CLEAR_DEPTH] );
			OUT_RING( 0x0 );

			ADVANCE_RING();
		}
	}

	/* Increment the clear counter.  The client-side 3D driver must
	 * wait on this value before performing the clear ioctl.  We
	 * need this because the card's so damned fast...
	 */
	dev_priv->sarea_priv->last_clear++;

	BEGIN_RING( 4 );

	RADEON_CLEAR_AGE( dev_priv->sarea_priv->last_clear );
	RADEON_WAIT_UNTIL_IDLE();

	ADVANCE_RING();
}

static void radeon_cp_dispatch_swap( drm_device_t *dev )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_radeon_sarea_t *sarea_priv = dev_priv->sarea_priv;
	int nbox = sarea_priv->nbox;
	drm_clip_rect_t *pbox = sarea_priv->boxes;
	int i;
	RING_LOCALS;
	DRM_DEBUG( "\n" );

	/* Do some trivial performance monitoring...
	 */
	if (dev_priv->do_boxes)
		radeon_cp_performance_boxes( dev_priv );


	/* Wait for the 3D stream to idle before dispatching the bitblt.
	 * This will prevent data corruption between the two streams.
	 */
	BEGIN_RING( 2 );

	RADEON_WAIT_UNTIL_3D_IDLE();

	ADVANCE_RING();

	for ( i = 0 ; i < nbox ; i++ ) {
		int x = pbox[i].x1;
		int y = pbox[i].y1;
		int w = pbox[i].x2 - x;
		int h = pbox[i].y2 - y;

		DRM_DEBUG( "dispatch swap %d,%d-%d,%d\n",
			   x, y, w, h );

		BEGIN_RING( 7 );

		OUT_RING( CP_PACKET3( RADEON_CNTL_BITBLT_MULTI, 5 ) );
		OUT_RING( RADEON_GMC_SRC_PITCH_OFFSET_CNTL |
			  RADEON_GMC_DST_PITCH_OFFSET_CNTL |
			  RADEON_GMC_BRUSH_NONE |
			  (dev_priv->color_fmt << 8) |
			  RADEON_GMC_SRC_DATATYPE_COLOR |
			  RADEON_ROP3_S |
			  RADEON_DP_SRC_SOURCE_MEMORY |
			  RADEON_GMC_CLR_CMP_CNTL_DIS |
			  RADEON_GMC_WR_MSK_DIS );
		
		/* Make this work even if front & back are flipped:
		 */
		if (dev_priv->current_page == 0) {
			OUT_RING( dev_priv->back_pitch_offset );
			OUT_RING( dev_priv->front_pitch_offset );
		} 
		else {
			OUT_RING( dev_priv->front_pitch_offset );
			OUT_RING( dev_priv->back_pitch_offset );
		}

		OUT_RING( (x << 16) | y );
		OUT_RING( (x << 16) | y );
		OUT_RING( (w << 16) | h );

		ADVANCE_RING();
	}

	/* Increment the frame counter.  The client-side 3D driver must
	 * throttle the framerate by waiting for this value before
	 * performing the swapbuffer ioctl.
	 */
	dev_priv->sarea_priv->last_frame++;

	BEGIN_RING( 4 );

	RADEON_FRAME_AGE( dev_priv->sarea_priv->last_frame );
	RADEON_WAIT_UNTIL_2D_IDLE();

	ADVANCE_RING();
}

static void radeon_cp_dispatch_flip( drm_device_t *dev )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_sarea_t *sarea = (drm_sarea_t *)dev_priv->sarea->handle;
	int offset = (dev_priv->current_page == 1)
		   ? dev_priv->front_offset : dev_priv->back_offset;
	RING_LOCALS;
	DRM_DEBUG( "%s: page=%d pfCurrentPage=%d\n", 
		__FUNCTION__, 
		dev_priv->current_page,
		dev_priv->sarea_priv->pfCurrentPage);

	/* Do some trivial performance monitoring...
	 */
	if (dev_priv->do_boxes) {
		dev_priv->stats.boxes |= RADEON_BOX_FLIP;
		radeon_cp_performance_boxes( dev_priv );
	}

	/* Update the frame offsets for both CRTCs
	 */
	BEGIN_RING( 6 );

	RADEON_WAIT_UNTIL_3D_IDLE();
	OUT_RING_REG( RADEON_CRTC_OFFSET, ( ( sarea->frame.y * dev_priv->front_pitch
					      + sarea->frame.x 
					      * ( dev_priv->color_fmt - 2 ) ) & ~7 )
					  + offset );
	OUT_RING_REG( RADEON_CRTC2_OFFSET, dev_priv->sarea_priv->crtc2_base
					   + offset );

	ADVANCE_RING();

	/* Increment the frame counter.  The client-side 3D driver must
	 * throttle the framerate by waiting for this value before
	 * performing the swapbuffer ioctl.
	 */
	dev_priv->sarea_priv->last_frame++;
	dev_priv->sarea_priv->pfCurrentPage = dev_priv->current_page =
					      1 - dev_priv->current_page;

	BEGIN_RING( 2 );

	RADEON_FRAME_AGE( dev_priv->sarea_priv->last_frame );

	ADVANCE_RING();
}

static int bad_prim_vertex_nr( int primitive, int nr )
{
	switch (primitive & RADEON_PRIM_TYPE_MASK) {
	case RADEON_PRIM_TYPE_NONE:
	case RADEON_PRIM_TYPE_POINT:
		return nr < 1;
	case RADEON_PRIM_TYPE_LINE:
		return (nr & 1) || nr == 0;
	case RADEON_PRIM_TYPE_LINE_STRIP:
		return nr < 2;
	case RADEON_PRIM_TYPE_TRI_LIST:
	case RADEON_PRIM_TYPE_3VRT_POINT_LIST:
	case RADEON_PRIM_TYPE_3VRT_LINE_LIST:
	case RADEON_PRIM_TYPE_RECT_LIST:
		return nr % 3 || nr == 0;
	case RADEON_PRIM_TYPE_TRI_FAN:
	case RADEON_PRIM_TYPE_TRI_STRIP:
		return nr < 3;
	default:
		return 1;
	}	
}



typedef struct {
	unsigned int start;
	unsigned int finish;
	unsigned int prim;
	unsigned int numverts;
	unsigned int offset;   
        unsigned int vc_format;
} drm_radeon_tcl_prim_t;

static void radeon_cp_dispatch_vertex( drm_device_t *dev,
				       drm_buf_t *buf,
				       drm_radeon_tcl_prim_t *prim )

{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_radeon_sarea_t *sarea_priv = dev_priv->sarea_priv;
	int offset = dev_priv->gart_buffers_offset + buf->offset + prim->start;
	int numverts = (int)prim->numverts;
	int nbox = sarea_priv->nbox;
	int i = 0;
	RING_LOCALS;

	DRM_DEBUG("hwprim 0x%x vfmt 0x%x %d..%d %d verts\n",
		  prim->prim,
		  prim->vc_format,
		  prim->start,
		  prim->finish,
		  prim->numverts);

	if (bad_prim_vertex_nr( prim->prim, prim->numverts )) {
		DRM_ERROR( "bad prim %x numverts %d\n", 
			   prim->prim, prim->numverts );
		return;
	}

	do {
		/* Emit the next cliprect */
		if ( i < nbox ) {
			radeon_emit_clip_rect( dev_priv, 
					       &sarea_priv->boxes[i] );
		}

		/* Emit the vertex buffer rendering commands */
		BEGIN_RING( 5 );

		OUT_RING( CP_PACKET3( RADEON_3D_RNDR_GEN_INDX_PRIM, 3 ) );
		OUT_RING( offset );
		OUT_RING( numverts );
		OUT_RING( prim->vc_format );
		OUT_RING( prim->prim | RADEON_PRIM_WALK_LIST |
			  RADEON_COLOR_ORDER_RGBA |
			  RADEON_VTX_FMT_RADEON_MODE |
			  (numverts << RADEON_NUM_VERTICES_SHIFT) );

		ADVANCE_RING();

		i++;
	} while ( i < nbox );
}



static void radeon_cp_discard_buffer( drm_device_t *dev, drm_buf_t *buf )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_radeon_buf_priv_t *buf_priv = buf->dev_private;
	RING_LOCALS;

	buf_priv->age = ++dev_priv->sarea_priv->last_dispatch;

	/* Emit the vertex buffer age */
	BEGIN_RING( 2 );
	RADEON_DISPATCH_AGE( buf_priv->age );
	ADVANCE_RING();

	buf->pending = 1;
	buf->used = 0;
}

static void radeon_cp_dispatch_indirect( drm_device_t *dev,
					 drm_buf_t *buf,
					 int start, int end )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	RING_LOCALS;
	DRM_DEBUG( "indirect: buf=%d s=0x%x e=0x%x\n",
		   buf->idx, start, end );

	if ( start != end ) {
		int offset = (dev_priv->gart_buffers_offset
			      + buf->offset + start);
		int dwords = (end - start + 3) / sizeof(u32);

		/* Indirect buffer data must be an even number of
		 * dwords, so if we've been given an odd number we must
		 * pad the data with a Type-2 CP packet.
		 */
		if ( dwords & 1 ) {
			u32 *data = (u32 *)
				((char *)dev->agp_buffer_map->handle
				 + buf->offset + start);
			data[dwords++] = RADEON_CP_PACKET2;
		}

		/* Fire off the indirect buffer */
		BEGIN_RING( 3 );

		OUT_RING( CP_PACKET0( RADEON_CP_IB_BASE, 1 ) );
		OUT_RING( offset );
		OUT_RING( dwords );

		ADVANCE_RING();
	}
}


static void radeon_cp_dispatch_indices( drm_device_t *dev,
					drm_buf_t *elt_buf,
					drm_radeon_tcl_prim_t *prim )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_radeon_sarea_t *sarea_priv = dev_priv->sarea_priv;
	int offset = dev_priv->gart_buffers_offset + prim->offset;
	u32 *data;
	int dwords;
	int i = 0;
	int start = prim->start + RADEON_INDEX_PRIM_OFFSET;
	int count = (prim->finish - start) / sizeof(u16);
	int nbox = sarea_priv->nbox;

	DRM_DEBUG("hwprim 0x%x vfmt 0x%x %d..%d offset: %x nr %d\n",
		  prim->prim,
		  prim->vc_format,
		  prim->start,
		  prim->finish,
		  prim->offset,
		  prim->numverts);

	if (bad_prim_vertex_nr( prim->prim, count )) {
		DRM_ERROR( "bad prim %x count %d\n", 
			   prim->prim, count );
		return;
	}


	if ( start >= prim->finish ||
	     (prim->start & 0x7) ) {
		DRM_ERROR( "buffer prim %d\n", prim->prim );
		return;
	}

	dwords = (prim->finish - prim->start + 3) / sizeof(u32);

	data = (u32 *)((char *)dev->agp_buffer_map->handle +
		       elt_buf->offset + prim->start);

	data[0] = CP_PACKET3( RADEON_3D_RNDR_GEN_INDX_PRIM, dwords-2 );
	data[1] = offset;
	data[2] = prim->numverts;
	data[3] = prim->vc_format;
	data[4] = (prim->prim |
		   RADEON_PRIM_WALK_IND |
		   RADEON_COLOR_ORDER_RGBA |
		   RADEON_VTX_FMT_RADEON_MODE |
		   (count << RADEON_NUM_VERTICES_SHIFT) );

	do {
		if ( i < nbox ) 
			radeon_emit_clip_rect( dev_priv, 
					       &sarea_priv->boxes[i] );

		radeon_cp_dispatch_indirect( dev, elt_buf,
					     prim->start,
					     prim->finish );

		i++;
	} while ( i < nbox );

}

#define RADEON_MAX_TEXTURE_SIZE (RADEON_BUFFER_SIZE - 8 * sizeof(u32))

static int radeon_cp_dispatch_texture( DRMFILE filp,
				       drm_device_t *dev,
				       drm_radeon_texture_t *tex,
				       drm_radeon_tex_image_t *image )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_file_t *filp_priv;
	drm_buf_t *buf;
	u32 format;
	u32 *buffer;
	const u8 __user *data;
	int size, dwords, tex_width, blit_width;
	u32 height;
	int i;
	RING_LOCALS;

	DRM_GET_PRIV_WITH_RETURN( filp_priv, filp );

	if ( radeon_check_and_fixup_offset( dev_priv, filp_priv, &tex->offset ) ) {
		DRM_ERROR( "Invalid destination offset\n" );
		return DRM_ERR( EINVAL );
	}

	dev_priv->stats.boxes |= RADEON_BOX_TEXTURE_LOAD;

	/* Flush the pixel cache.  This ensures no pixel data gets mixed
	 * up with the texture data from the host data blit, otherwise
	 * part of the texture image may be corrupted.
	 */
	BEGIN_RING( 4 );
	RADEON_FLUSH_CACHE();
	RADEON_WAIT_UNTIL_IDLE();
	ADVANCE_RING();

#ifdef __BIG_ENDIAN
	/* The Mesa texture functions provide the data in little endian as the
	 * chip wants it, but we need to compensate for the fact that the CP
	 * ring gets byte-swapped
	 */
	BEGIN_RING( 2 );
	OUT_RING_REG( RADEON_RBBM_GUICNTL, RADEON_HOST_DATA_SWAP_32BIT );
	ADVANCE_RING();
#endif


	/* The compiler won't optimize away a division by a variable,
	 * even if the only legal values are powers of two.  Thus, we'll
	 * use a shift instead.
	 */
	switch ( tex->format ) {
	case RADEON_TXFORMAT_ARGB8888:
	case RADEON_TXFORMAT_RGBA8888:
		format = RADEON_COLOR_FORMAT_ARGB8888;
		tex_width = tex->width * 4;
		blit_width = image->width * 4;
		break;
	case RADEON_TXFORMAT_AI88:
	case RADEON_TXFORMAT_ARGB1555:
	case RADEON_TXFORMAT_RGB565:
	case RADEON_TXFORMAT_ARGB4444:
	case RADEON_TXFORMAT_VYUY422:
	case RADEON_TXFORMAT_YVYU422:
		format = RADEON_COLOR_FORMAT_RGB565;
		tex_width = tex->width * 2;
		blit_width = image->width * 2;
		break;
	case RADEON_TXFORMAT_I8:
	case RADEON_TXFORMAT_RGB332:
		format = RADEON_COLOR_FORMAT_CI8;
		tex_width = tex->width * 1;
		blit_width = image->width * 1;
		break;
	default:
		DRM_ERROR( "invalid texture format %d\n", tex->format );
		return DRM_ERR(EINVAL);
	}

	DRM_DEBUG("tex=%dx%d blit=%d\n", tex_width, tex->height, blit_width );

	do {
		DRM_DEBUG( "tex: ofs=0x%x p=%d f=%d x=%hd y=%hd w=%hd h=%hd\n",
			   tex->offset >> 10, tex->pitch, tex->format,
			   image->x, image->y, image->width, image->height );

		/* Make a copy of some parameters in case we have to
		 * update them for a multi-pass texture blit.
		 */
		height = image->height;
		data = (const u8 __user *)image->data;
		
		size = height * blit_width;

		if ( size > RADEON_MAX_TEXTURE_SIZE ) {
			height = RADEON_MAX_TEXTURE_SIZE / blit_width;
			size = height * blit_width;
		} else if ( size < 4 && size > 0 ) {
			size = 4;
		} else if ( size == 0 ) {
			return 0;
		}

		buf = radeon_freelist_get( dev );
		if ( 0 && !buf ) {
			radeon_do_cp_idle( dev_priv );
			buf = radeon_freelist_get( dev );
		}
		if ( !buf ) {
			DRM_DEBUG("radeon_cp_dispatch_texture: EAGAIN\n");
			if (DRM_COPY_TO_USER( tex->image, image, sizeof(*image) ))
				return DRM_ERR(EFAULT);
			return DRM_ERR(EAGAIN);
		}


		/* Dispatch the indirect buffer.
		 */
		buffer = (u32*)((char*)dev->agp_buffer_map->handle + buf->offset);
		dwords = size / 4;
		buffer[0] = CP_PACKET3( RADEON_CNTL_HOSTDATA_BLT, dwords + 6 );
		buffer[1] = (RADEON_GMC_DST_PITCH_OFFSET_CNTL |
			     RADEON_GMC_BRUSH_NONE |
			     (format << 8) |
			     RADEON_GMC_SRC_DATATYPE_COLOR |
			     RADEON_ROP3_S |
			     RADEON_DP_SRC_SOURCE_HOST_DATA |
			     RADEON_GMC_CLR_CMP_CNTL_DIS |
			     RADEON_GMC_WR_MSK_DIS);
		
		buffer[2] = (tex->pitch << 22) | (tex->offset >> 10);
		buffer[3] = 0xffffffff;
		buffer[4] = 0xffffffff;
		buffer[5] = (image->y << 16) | image->x;
		buffer[6] = (height << 16) | image->width;
		buffer[7] = dwords;
		buffer += 8;

		if ( tex_width >= 32 ) {
			/* Texture image width is larger than the minimum, so we
			 * can upload it directly.
			 */
			if ( DRM_COPY_FROM_USER( buffer, data, 
						 dwords * sizeof(u32) ) ) {
				DRM_ERROR( "EFAULT on data, %d dwords\n", 
					   dwords );
				return DRM_ERR(EFAULT);
			}
		} else {
			/* Texture image width is less than the minimum, so we
			 * need to pad out each image scanline to the minimum
			 * width.
			 */
			for ( i = 0 ; i < tex->height ; i++ ) {
				if ( DRM_COPY_FROM_USER( buffer, data, 
							 tex_width ) ) {
					DRM_ERROR( "EFAULT on pad, %d bytes\n",
						   tex_width );
					return DRM_ERR(EFAULT);
				}
				buffer += 8;
				data += tex_width;
			}
		}

		buf->filp = filp;
		buf->used = (dwords + 8) * sizeof(u32);
		radeon_cp_dispatch_indirect( dev, buf, 0, buf->used );
		radeon_cp_discard_buffer( dev, buf );

		/* Update the input parameters for next time */
		image->y += height;
		image->height -= height;
		image->data = (const u8 __user *)image->data + size;
	} while (image->height > 0);

	/* Flush the pixel cache after the blit completes.  This ensures
	 * the texture data is written out to memory before rendering
	 * continues.
	 */
	BEGIN_RING( 4 );
	RADEON_FLUSH_CACHE();
	RADEON_WAIT_UNTIL_2D_IDLE();
	ADVANCE_RING();
	return 0;
}


static void radeon_cp_dispatch_stipple( drm_device_t *dev, u32 *stipple )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	int i;
	RING_LOCALS;
	DRM_DEBUG( "\n" );

	BEGIN_RING( 35 );

	OUT_RING( CP_PACKET0( RADEON_RE_STIPPLE_ADDR, 0 ) );
	OUT_RING( 0x00000000 );

	OUT_RING( CP_PACKET0_TABLE( RADEON_RE_STIPPLE_DATA, 31 ) );
	for ( i = 0 ; i < 32 ; i++ ) {
		OUT_RING( stipple[i] );
	}

	ADVANCE_RING();
}


/* ================================================================
 * IOCTL functions
 */

int radeon_cp_clear( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_radeon_sarea_t *sarea_priv = dev_priv->sarea_priv;
	drm_radeon_clear_t clear;
	drm_radeon_clear_rect_t depth_boxes[RADEON_NR_SAREA_CLIPRECTS];
	DRM_DEBUG( "\n" );

	LOCK_TEST_WITH_RETURN( dev, filp );

	DRM_COPY_FROM_USER_IOCTL( clear, (drm_radeon_clear_t __user *)data,
			     sizeof(clear) );

	RING_SPACE_TEST_WITH_RETURN( dev_priv );

	if ( sarea_priv->nbox > RADEON_NR_SAREA_CLIPRECTS )
		sarea_priv->nbox = RADEON_NR_SAREA_CLIPRECTS;

	if ( DRM_COPY_FROM_USER( &depth_boxes, clear.depth_boxes,
			     sarea_priv->nbox * sizeof(depth_boxes[0]) ) )
		return DRM_ERR(EFAULT);

	radeon_cp_dispatch_clear( dev, &clear, depth_boxes );

	COMMIT_RING();
	return 0;
}


/* Not sure why this isn't set all the time:
 */ 
static int radeon_do_init_pageflip( drm_device_t *dev )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	RING_LOCALS;

	DRM_DEBUG( "\n" );

	BEGIN_RING( 6 );
	RADEON_WAIT_UNTIL_3D_IDLE();
	OUT_RING( CP_PACKET0( RADEON_CRTC_OFFSET_CNTL, 0 ) );
	OUT_RING( RADEON_READ( RADEON_CRTC_OFFSET_CNTL ) | RADEON_CRTC_OFFSET_FLIP_CNTL );
	OUT_RING( CP_PACKET0( RADEON_CRTC2_OFFSET_CNTL, 0 ) );
	OUT_RING( RADEON_READ( RADEON_CRTC2_OFFSET_CNTL ) | RADEON_CRTC_OFFSET_FLIP_CNTL );
	ADVANCE_RING();

	dev_priv->page_flipping = 1;
	dev_priv->current_page = 0;
	dev_priv->sarea_priv->pfCurrentPage = dev_priv->current_page;

	return 0;
}

/* Called whenever a client dies, from DRM(release).
 * NOTE:  Lock isn't necessarily held when this is called!
 */
int radeon_do_cleanup_pageflip( drm_device_t *dev )
{
	drm_radeon_private_t *dev_priv = dev->dev_private;
	DRM_DEBUG( "\n" );

	if (dev_priv->current_page != 0)
		radeon_cp_dispatch_flip( dev );

	dev_priv->page_flipping = 0;
	return 0;
}

/* Swapping and flipping are different operations, need different ioctls.
 * They can & should be intermixed to support multiple 3d windows.  
 */
int radeon_cp_flip( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_radeon_private_t *dev_priv = dev->dev_private;
	DRM_DEBUG( "\n" );

	LOCK_TEST_WITH_RETURN( dev, filp );

	RING_SPACE_TEST_WITH_RETURN( dev_priv );

	if (!dev_priv->page_flipping) 
		radeon_do_init_pageflip( dev );
		
	radeon_cp_dispatch_flip( dev );

	COMMIT_RING();
	return 0;
}

int radeon_cp_swap( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_radeon_sarea_t *sarea_priv = dev_priv->sarea_priv;
	DRM_DEBUG( "\n" );

	LOCK_TEST_WITH_RETURN( dev, filp );

	RING_SPACE_TEST_WITH_RETURN( dev_priv );

	if ( sarea_priv->nbox > RADEON_NR_SAREA_CLIPRECTS )
		sarea_priv->nbox = RADEON_NR_SAREA_CLIPRECTS;

	radeon_cp_dispatch_swap( dev );
	dev_priv->sarea_priv->ctx_owner = 0;

	COMMIT_RING();
	return 0;
}

int radeon_cp_vertex( DRM_IOCTL_ARGS )
{
	DRM_DEVICE;
	drm_radeon_private_t *dev_priv = dev->dev_private;
	drm_file_t *filp_priv;
	drm_radeon_sarea_t *sarea_priv = dev_priv->sarea_priv;
	drm_device_dma_t *dma = dev->dma;
	drm_buf_t *buf;
	drm_radeon_vertex_t vertex;
	drm_radeon_tcl_prim_t prim;

	LOCK_TEST_WITH_RETURN( dev, filp );

	if ( !dev_priv ) {