aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/gpu/drm/amd/amdgpu/atom.c
blob: 4cfc786699c7fcaac2b903c463ea85bbd9379812 (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
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494

  int processed = 0;

  if (ctx->passphrase_cb)
    {
      err = _gpgme_passphrase_command_handler (ctx, status, args,
					       fd, &processed);
      if (err)
	return err;
    }

  if (!processed)
    {
      void *hook;
      op_data_t opd;

      err = _gpgme_op_data_lookup (ctx, OPDATA_EDIT, &hook, -1, NULL);
      opd = hook;
      if (err)
	return err;

      /* FIXME: We expect the user to handle _all_ status codes.
	 Later, we may fix the callback interface to allow the user
	 indicate if it processed the status code or not.  */
      *processed_r = 1;

      return (*opd->fnc) (opd->fnc_value, status, args, fd);
    }

  *processed_r = processed;
  return 0;
}


static gpgme_error_t
edit_start (gpgme_ctx_t ctx, int synchronous, int type, gpgme_key_t key,
	    gpgme_edit_cb_t fnc, void *fnc_value, gpgme_data_t out)
{
  gpgme_error_t err;
  void *hook;
  op_data_t opd;

  err = _gpgme_op_reset (ctx, synchronous);
  if (err)
    return err;

  if (!fnc || !out)
    return gpg_error (GPG_ERR_INV_VALUE);

  err = _gpgme_op_data_lookup (ctx, OPDATA_EDIT, &hook, sizeof (*opd), NULL);
  opd = hook;
  if (err)
    return err;

  opd->fnc = fnc;
  opd->fnc_value = fnc_value;

  err = _gpgme_engine_set_command_handler (ctx->engine, command_handler,
					   ctx, out);
  if (err)
    return err;

  _gpgme_engine_set_status_handler (ctx->engine, edit_status_handler, ctx);

  return _gpgme_engine_op_edit (ctx->engine, type, key, out, ctx);
}


gpgme_error_t
gpgme_op_edit_start (gpgme_ctx_t ctx, gpgme_key_t key,
		     gpgme_edit_cb_t fnc, void *fnc_value, gpgme_data_t out)
{
  gpgme_error_t err;

  TRACE_BEG5 (DEBUG_CTX, "gpgme_op_edit_start", ctx,
	      "key=%p (%s), fnc=%p fnc_value=%p, out=%p", key,
	      (key && key->subkeys && key->subkeys->fpr) ?
	      key->subkeys->fpr : "invalid", fnc, fnc_value, out);

  if (!ctx)
    return TRACE_ERR (gpg_error (GPG_ERR_INV_VALUE));

  err = edit_start (ctx, 0, 0, key, fnc, fnc_value, out);
  return err;
}


/* Edit the key KEY.  Send status and command requests to FNC and
   output of edit commands to OUT.  */
gpgme_error_t
gpgme_op_edit (gpgme_ctx_t ctx, gpgme_key_t key,
	       gpgme_edit_cb_t fnc, void *fnc_value, gpgme_data_t out)
{
  gpgme_error_t err;

  TRACE_BEG5 (DEBUG_CTX, "gpgme_op_edit", ctx,
	      "key=%p (%s), fnc=%p fnc_value=%p, out=%p", key,
	      (key && key->subkeys && key->subkeys->fpr) ?
	      key->subkeys->fpr : "invalid", fnc, fnc_value, out);

  if (!ctx)
    return TRACE_ERR (gpg_error (GPG_ERR_INV_VALUE));

  err = edit_start (ctx, 1, 0, key, fnc, fnc_value, out);

  if (!err)
    err = _gpgme_wait_one (ctx);
  return TRACE_ERR (err);
}


gpgme_error_t
gpgme_op_card_edit_start (gpgme_ctx_t ctx, gpgme_key_t key,
			  gpgme_edit_cb_t fnc, void *fnc_value,
			  gpgme_data_t out)
{
  gpgme_error_t err;

  TRACE_BEG5 (DEBUG_CTX, "gpgme_op_card_edit_start", ctx,
	      "key=%p (%s), fnc=%p fnc_value=%p, out=%p", key,
	      (key && key->subkeys && key->subkeys->fpr) ?
	      key->subkeys->fpr : "invalid", fnc, fnc_value, out);

  if (!ctx)
    return TRACE_ERR (gpg_error (GPG_ERR_INV_VALUE));

  err = edit_start (ctx, 0, 1, key, fnc, fnc_value, out);
  return err;
}


/* Edit the card for the key KEY.  Send status and command requests to
   FNC and output of edit commands to OUT.  */
gpgme_error_t
gpgme_op_card_edit (gpgme_ctx_t ctx, gpgme_key_t key,
		    gpgme_edit_cb_t fnc, void *fnc_value, gpgme_data_t out)
{
  gpgme_error_t err;

  TRACE_BEG5 (DEBUG_CTX, "gpgme_op_card_edit", ctx,
	      "key=%p (%s), fnc=%p fnc_value=%p, out=%p", key,
	      (key && key->subkeys && key->subkeys->fpr) ?
	      key->subkeys->fpr : "invalid", fnc, fnc_value, out);

  if (!ctx)
    return TRACE_ERR (gpg_error (GPG_ERR_INV_VALUE));

  err = edit_start (ctx, 1, 1, key, fnc, fnc_value, out);
  if (!err)
    err = _gpgme_wait_one (ctx);
  return TRACE_ERR (err);
}
1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427
/*
 * Copyright 2008 Advanced Micro Devices, Inc.
 *
 * 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 shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
 * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) 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.
 *
 * Author: Stanislaw Skowronek
 */

#include <linux/module.h>
#include <linux/sched.h>
#include <linux/slab.h>
#include <asm/unaligned.h>

#include <drm/drm_util.h>

#define ATOM_DEBUG

#include "atom.h"
#include "atom-names.h"
#include "atom-bits.h"
#include "amdgpu.h"

#define ATOM_COND_ABOVE		0
#define ATOM_COND_ABOVEOREQUAL	1
#define ATOM_COND_ALWAYS	2
#define ATOM_COND_BELOW		3
#define ATOM_COND_BELOWOREQUAL	4
#define ATOM_COND_EQUAL		5
#define ATOM_COND_NOTEQUAL	6

#define ATOM_PORT_ATI	0
#define ATOM_PORT_PCI	1
#define ATOM_PORT_SYSIO	2

#define ATOM_UNIT_MICROSEC	0
#define ATOM_UNIT_MILLISEC	1

#define PLL_INDEX	2
#define PLL_DATA	3

#define ATOM_CMD_TIMEOUT_SEC	20

typedef struct {
	struct atom_context *ctx;
	uint32_t *ps, *ws;
	int ps_shift;
	uint16_t start;
	unsigned last_jump;
	unsigned long last_jump_jiffies;
	bool abort;
} atom_exec_context;

int amdgpu_atom_debug = 0;
static int amdgpu_atom_execute_table_locked(struct atom_context *ctx, int index, uint32_t * params);
int amdgpu_atom_execute_table(struct atom_context *ctx, int index, uint32_t * params);

static uint32_t atom_arg_mask[8] =
    { 0xFFFFFFFF, 0xFFFF, 0xFFFF00, 0xFFFF0000, 0xFF, 0xFF00, 0xFF0000,
0xFF000000 };
static int atom_arg_shift[8] = { 0, 0, 8, 16, 0, 8, 16, 24 };

static int atom_dst_to_src[8][4] = {
	/* translate destination alignment field to the source alignment encoding */
	{0, 0, 0, 0},
	{1, 2, 3, 0},
	{1, 2, 3, 0},
	{1, 2, 3, 0},
	{4, 5, 6, 7},
	{4, 5, 6, 7},
	{4, 5, 6, 7},
	{4, 5, 6, 7},
};
static int atom_def_dst[8] = { 0, 0, 1, 2, 0, 1, 2, 3 };

static int debug_depth = 0;
#ifdef ATOM_DEBUG
static void debug_print_spaces(int n)
{
	while (n--)
		printk("   ");
}

#define DEBUG(...) do if (amdgpu_atom_debug) { printk(KERN_DEBUG __VA_ARGS__); } while (0)
#define SDEBUG(...) do if (amdgpu_atom_debug) { printk(KERN_DEBUG); debug_print_spaces(debug_depth); printk(__VA_ARGS__); } while (0)
#else
#define DEBUG(...) do { } while (0)
#define SDEBUG(...) do { } while (0)
#endif

static uint32_t atom_iio_execute(struct atom_context *ctx, int base,
				 uint32_t index, uint32_t data)
{
	uint32_t temp = 0xCDCDCDCD;

	while (1)
		switch (CU8(base)) {
		case ATOM_IIO_NOP:
			base++;
			break;
		case ATOM_IIO_READ:
			temp = ctx->card->ioreg_read(ctx->card, CU16(base + 1));
			base += 3;
			break;
		case ATOM_IIO_WRITE:
			ctx->card->ioreg_write(ctx->card, CU16(base + 1), temp);
			base += 3;
			break;
		case ATOM_IIO_CLEAR:
			temp &=
			    ~((0xFFFFFFFF >> (32 - CU8(base + 1))) <<
			      CU8(base + 2));
			base += 3;
			break;
		case ATOM_IIO_SET:
			temp |=
			    (0xFFFFFFFF >> (32 - CU8(base + 1))) << CU8(base +
									2);
			base += 3;
			break;
		case ATOM_IIO_MOVE_INDEX:
			temp &=
			    ~((0xFFFFFFFF >> (32 - CU8(base + 1))) <<
			      CU8(base + 3));
			temp |=
			    ((index >> CU8(base + 2)) &
			     (0xFFFFFFFF >> (32 - CU8(base + 1)))) << CU8(base +
									  3);
			base += 4;
			break;
		case ATOM_IIO_MOVE_DATA:
			temp &=
			    ~((0xFFFFFFFF >> (32 - CU8(base + 1))) <<
			      CU8(base + 3));
			temp |=
			    ((data >> CU8(base + 2)) &
			     (0xFFFFFFFF >> (32 - CU8(base + 1)))) << CU8(base +
									  3);
			base += 4;
			break;
		case ATOM_IIO_MOVE_ATTR:
			temp &=
			    ~((0xFFFFFFFF >> (32 - CU8(base + 1))) <<
			      CU8(base + 3));
			temp |=
			    ((ctx->
			      io_attr >> CU8(base + 2)) & (0xFFFFFFFF >> (32 -
									  CU8
									  (base
									   +
									   1))))
			    << CU8(base + 3);
			base += 4;
			break;
		case ATOM_IIO_END:
			return temp;
		default:
			pr_info("Unknown IIO opcode\n");
			return 0;
		}
}

static uint32_t atom_get_src_int(atom_exec_context *ctx, uint8_t attr,
				 int *ptr, uint32_t *saved, int print)
{
	uint32_t idx, val = 0xCDCDCDCD, align, arg;
	struct atom_context *gctx = ctx->ctx;
	arg = attr & 7;
	align = (attr >> 3) & 7;
	switch (arg) {
	case ATOM_ARG_REG:
		idx = U16(*ptr);
		(*ptr) += 2;
		if (print)
			DEBUG("REG[0x%04X]", idx);
		idx += gctx->reg_block;
		switch (gctx->io_mode) {
		case ATOM_IO_MM:
			val = gctx->card->reg_read(gctx->card, idx);
			break;
		case ATOM_IO_PCI:
			pr_info("PCI registers are not implemented\n");
			return 0;
		case ATOM_IO_SYSIO:
			pr_info("SYSIO registers are not implemented\n");
			return 0;
		default:
			if (!(gctx->io_mode & 0x80)) {
				pr_info("Bad IO mode\n");
				return 0;
			}
			if (!gctx->iio[gctx->io_mode & 0x7F]) {
				pr_info("Undefined indirect IO read method %d\n",
					gctx->io_mode & 0x7F);
				return 0;
			}
			val =
			    atom_iio_execute(gctx,
					     gctx->iio[gctx->io_mode & 0x7F],
					     idx, 0);
		}
		break;
	case ATOM_ARG_PS:
		idx = U8(*ptr);
		(*ptr)++;
		/* get_unaligned_le32 avoids unaligned accesses from atombios
		 * tables, noticed on a DEC Alpha. */
		val = get_unaligned_le32((u32 *)&ctx->ps[idx]);
		if (print)
			DEBUG("PS[0x%02X,0x%04X]", idx, val);
		break;
	case ATOM_ARG_WS:
		idx = U8(*ptr);
		(*ptr)++;
		if (print)
			DEBUG("WS[0x%02X]", idx);
		switch (idx) {
		case ATOM_WS_QUOTIENT:
			val = gctx->divmul[0];
			break;
		case ATOM_WS_REMAINDER:
			val = gctx->divmul[1];
			break;
		case ATOM_WS_DATAPTR:
			val = gctx->data_block;
			break;
		case ATOM_WS_SHIFT:
			val = gctx->shift;
			break;
		case ATOM_WS_OR_MASK:
			val = 1 << gctx->shift;
			break;
		case ATOM_WS_AND_MASK:
			val = ~(1 << gctx->shift);
			break;
		case ATOM_WS_FB_WINDOW:
			val = gctx->fb_base;
			break;
		case ATOM_WS_ATTRIBUTES:
			val = gctx->io_attr;
			break;
		case ATOM_WS_REGPTR:
			val = gctx->reg_block;
			break;
		default:
			val = ctx->ws[idx];
		}
		break;
	case ATOM_ARG_ID:
		idx = U16(*ptr);
		(*ptr) += 2;
		if (print) {
			if (gctx->data_block)
				DEBUG("ID[0x%04X+%04X]", idx, gctx->data_block);
			else
				DEBUG("ID[0x%04X]", idx);
		}
		val = U32(idx + gctx->data_block);
		break;
	case ATOM_ARG_FB:
		idx = U8(*ptr);
		(*ptr)++;
		if ((gctx->fb_base + (idx * 4)) > gctx->scratch_size_bytes) {
			DRM_ERROR("ATOM: fb read beyond scratch region: %d vs. %d\n",
				  gctx->fb_base + (idx * 4), gctx->scratch_size_bytes);
			val = 0;
		} else
			val = gctx->scratch[(gctx->fb_base / 4) + idx];
		if (print)
			DEBUG("FB[0x%02X]", idx);
		break;
	case ATOM_ARG_IMM:
		switch (align) {
		case ATOM_SRC_DWORD:
			val = U32(*ptr);
			(*ptr) += 4;
			if (print)
				DEBUG("IMM 0x%08X\n", val);
			return val;
		case ATOM_SRC_WORD0:
		case ATOM_SRC_WORD8:
		case ATOM_SRC_WORD16:
			val = U16(*ptr);
			(*ptr) += 2;
			if (print)
				DEBUG("IMM 0x%04X\n", val);
			return val;
		case ATOM_SRC_BYTE0:
		case ATOM_SRC_BYTE8:
		case ATOM_SRC_BYTE16:
		case ATOM_SRC_BYTE24:
			val = U8(*ptr);
			(*ptr)++;
			if (print)
				DEBUG("IMM 0x%02X\n", val);
			return val;
		}
		return 0;
	case ATOM_ARG_PLL:
		idx = U8(*ptr);
		(*ptr)++;
		if (print)
			DEBUG("PLL[0x%02X]", idx);
		val = gctx->card->pll_read(gctx->card, idx);
		break;
	case ATOM_ARG_MC:
		idx = U8(*ptr);
		(*ptr)++;
		if (print)
			DEBUG("MC[0x%02X]", idx);
		val = gctx->card->mc_read(gctx->card, idx);
		break;
	}
	if (saved)
		*saved = val;
	val &= atom_arg_mask[align];
	val >>= atom_arg_shift[align];
	if (print)
		switch (align) {
		case ATOM_SRC_DWORD:
			DEBUG(".[31:0] -> 0x%08X\n", val);
			break;
		case ATOM_SRC_WORD0:
			DEBUG(".[15:0] -> 0x%04X\n", val);
			break;
		case ATOM_SRC_WORD8:
			DEBUG(".[23:8] -> 0x%04X\n", val);
			break;
		case ATOM_SRC_WORD16:
			DEBUG(".[31:16] -> 0x%04X\n", val);
			break;
		case ATOM_SRC_BYTE0:
			DEBUG(".[7:0] -> 0x%02X\n", val);
			break;
		case ATOM_SRC_BYTE8:
			DEBUG(".[15:8] -> 0x%02X\n", val);
			break;
		case ATOM_SRC_BYTE16:
			DEBUG(".[23:16] -> 0x%02X\n", val);
			break;
		case ATOM_SRC_BYTE24:
			DEBUG(".[31:24] -> 0x%02X\n", val);
			break;
		}
	return val;
}

static void atom_skip_src_int(atom_exec_context *ctx, uint8_t attr, int *ptr)
{
	uint32_t align = (attr >> 3) & 7, arg = attr & 7;
	switch (arg) {
	case ATOM_ARG_REG:
	case ATOM_ARG_ID:
		(*ptr) += 2;
		break;
	case ATOM_ARG_PLL:
	case ATOM_ARG_MC:
	case ATOM_ARG_PS:
	case ATOM_ARG_WS:
	case ATOM_ARG_FB:
		(*ptr)++;
		break;
	case ATOM_ARG_IMM:
		switch (align) {
		case ATOM_SRC_DWORD:
			(*ptr) += 4;
			return;
		case ATOM_SRC_WORD0:
		case ATOM_SRC_WORD8:
		case ATOM_SRC_WORD16:
			(*ptr) += 2;
			return;
		case ATOM_SRC_BYTE0:
		case ATOM_SRC_BYTE8:
		case ATOM_SRC_BYTE16:
		case ATOM_SRC_BYTE24:
			(*ptr)++;
			return;
		}
		return;
	}
}

static uint32_t atom_get_src(atom_exec_context *ctx, uint8_t attr, int *ptr)
{
	return atom_get_src_int(ctx, attr, ptr, NULL, 1);
}

static uint32_t atom_get_src_direct(atom_exec_context *ctx, uint8_t align, int *ptr)
{
	uint32_t val = 0xCDCDCDCD;

	switch (align) {
	case ATOM_SRC_DWORD:
		val = U32(*ptr);
		(*ptr) += 4;
		break;
	case ATOM_SRC_WORD0:
	case ATOM_SRC_WORD8:
	case ATOM_SRC_WORD16:
		val = U16(*ptr);
		(*ptr) += 2;
		break;
	case ATOM_SRC_BYTE0:
	case ATOM_SRC_BYTE8:
	case ATOM_SRC_BYTE16:
	case ATOM_SRC_BYTE24:
		val = U8(*ptr);
		(*ptr)++;
		break;
	}
	return val;
}

static uint32_t atom_get_dst(atom_exec_context *ctx, int arg, uint8_t attr,
			     int *ptr, uint32_t *saved, int print)
{
	return atom_get_src_int(ctx,
				arg | atom_dst_to_src[(attr >> 3) &
						      7][(attr >> 6) & 3] << 3,
				ptr, saved, print);
}

static void atom_skip_dst(atom_exec_context *ctx, int arg, uint8_t attr, int *ptr)
{
	atom_skip_src_int(ctx,
			  arg | atom_dst_to_src[(attr >> 3) & 7][(attr >> 6) &
								 3] << 3, ptr);
}

static void atom_put_dst(atom_exec_context *ctx, int arg, uint8_t attr,
			 int *ptr, uint32_t val, uint32_t saved)
{
	uint32_t align =
	    atom_dst_to_src[(attr >> 3) & 7][(attr >> 6) & 3], old_val =
	    val, idx;
	struct atom_context *gctx = ctx->ctx;
	old_val &= atom_arg_mask[align] >> atom_arg_shift[align];
	val <<= atom_arg_shift[align];
	val &= atom_arg_mask[align];
	saved &= ~atom_arg_mask[align];
	val |= saved;
	switch (arg) {
	case ATOM_ARG_REG:
		idx = U16(*ptr);
		(*ptr) += 2;
		DEBUG("REG[0x%04X]", idx);
		idx += gctx->reg_block;
		switch (gctx->io_mode) {
		case ATOM_IO_MM:
			if (idx == 0)
				gctx->card->reg_write(gctx->card, idx,
						      val << 2);
			else
				gctx->card->reg_write(gctx->card, idx, val);
			break;
		case ATOM_IO_PCI:
			pr_info("PCI registers are not implemented\n");
			return;
		case ATOM_IO_SYSIO:
			pr_info("SYSIO registers are not implemented\n");
			return;
		default:
			if (!(gctx->io_mode & 0x80)) {
				pr_info("Bad IO mode\n");
				return;
			}
			if (!gctx->iio[gctx->io_mode & 0xFF]) {
				pr_info("Undefined indirect IO write method %d\n",
					gctx->io_mode & 0x7F);
				return;
			}
			atom_iio_execute(gctx, gctx->iio[gctx->io_mode & 0xFF],
					 idx, val);
		}
		break;
	case ATOM_ARG_PS:
		idx = U8(*ptr);
		(*ptr)++;
		DEBUG("PS[0x%02X]", idx);
		ctx->ps[idx] = cpu_to_le32(val);
		break;
	case ATOM_ARG_WS:
		idx = U8(*ptr);
		(*ptr)++;
		DEBUG("WS[0x%02X]", idx);
		switch (idx) {
		case ATOM_WS_QUOTIENT:
			gctx->divmul[0] = val;
			break;
		case ATOM_WS_REMAINDER:
			gctx->divmul[1] = val;
			break;
		case ATOM_WS_DATAPTR:
			gctx->data_block = val;
			break;
		case ATOM_WS_SHIFT:
			gctx->shift = val;
			break;
		case ATOM_WS_OR_MASK:
		case ATOM_WS_AND_MASK:
			break;
		case ATOM_WS_FB_WINDOW:
			gctx->fb_base = val;
			break;
		case ATOM_WS_ATTRIBUTES:
			gctx->io_attr = val;
			break;
		case ATOM_WS_REGPTR:
			gctx->reg_block = val;
			break;
		default:
			ctx->ws[idx] = val;
		}
		break;
	case ATOM_ARG_FB:
		idx = U8(*ptr);
		(*ptr)++;
		if ((gctx->fb_base + (idx * 4)) > gctx->scratch_size_bytes) {
			DRM_ERROR("ATOM: fb write beyond scratch region: %d vs. %d\n",
				  gctx->fb_base + (idx * 4), gctx->scratch_size_bytes);
		} else
			gctx->scratch[(gctx->fb_base / 4) + idx] = val;
		DEBUG("FB[0x%02X]", idx);
		break;
	case ATOM_ARG_PLL:
		idx = U8(*ptr);
		(*ptr)++;
		DEBUG("PLL[0x%02X]", idx);
		gctx->card->pll_write(gctx->card, idx, val);
		break;
	case ATOM_ARG_MC:
		idx = U8(*ptr);
		(*ptr)++;
		DEBUG("MC[0x%02X]", idx);
		gctx->card->mc_write(gctx->card, idx, val);
		return;
	}
	switch (align) {
	case ATOM_SRC_DWORD:
		DEBUG(".[31:0] <- 0x%08X\n", old_val);
		break;
	case ATOM_SRC_WORD0:
		DEBUG(".[15:0] <- 0x%04X\n", old_val);
		break;
	case ATOM_SRC_WORD8:
		DEBUG(".[23:8] <- 0x%04X\n", old_val);
		break;
	case ATOM_SRC_WORD16:
		DEBUG(".[31:16] <- 0x%04X\n", old_val);
		break;
	case ATOM_SRC_BYTE0:
		DEBUG(".[7:0] <- 0x%02X\n", old_val);
		break;
	case ATOM_SRC_BYTE8:
		DEBUG(".[15:8] <- 0x%02X\n", old_val);
		break;
	case ATOM_SRC_BYTE16:
		DEBUG(".[23:16] <- 0x%02X\n", old_val);
		break;
	case ATOM_SRC_BYTE24:
		DEBUG(".[31:24] <- 0x%02X\n", old_val);
		break;
	}
}

static void atom_op_add(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src, saved;
	int dptr = *ptr;
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	SDEBUG("   src: ");
	src = atom_get_src(ctx, attr, ptr);
	dst += src;
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_and(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src, saved;
	int dptr = *ptr;
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	SDEBUG("   src: ");
	src = atom_get_src(ctx, attr, ptr);
	dst &= src;
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_beep(atom_exec_context *ctx, int *ptr, int arg)
{
	printk("ATOM BIOS beeped!\n");
}

static void atom_op_calltable(atom_exec_context *ctx, int *ptr, int arg)
{
	int idx = U8((*ptr)++);
	int r = 0;

	if (idx < ATOM_TABLE_NAMES_CNT)
		SDEBUG("   table: %d (%s)\n", idx, atom_table_names[idx]);
	else
		SDEBUG("   table: %d\n", idx);
	if (U16(ctx->ctx->cmd_table + 4 + 2 * idx))
		r = amdgpu_atom_execute_table_locked(ctx->ctx, idx, ctx->ps + ctx->ps_shift);
	if (r) {
		ctx->abort = true;
	}
}

static void atom_op_clear(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t saved;
	int dptr = *ptr;
	attr &= 0x38;
	attr |= atom_def_dst[attr >> 3] << 6;
	atom_get_dst(ctx, arg, attr, ptr, &saved, 0);
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, 0, saved);
}

static void atom_op_compare(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src;
	SDEBUG("   src1: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, NULL, 1);
	SDEBUG("   src2: ");
	src = atom_get_src(ctx, attr, ptr);
	ctx->ctx->cs_equal = (dst == src);
	ctx->ctx->cs_above = (dst > src);
	SDEBUG("   result: %s %s\n", ctx->ctx->cs_equal ? "EQ" : "NE",
	       ctx->ctx->cs_above ? "GT" : "LE");
}

static void atom_op_delay(atom_exec_context *ctx, int *ptr, int arg)
{
	unsigned count = U8((*ptr)++);
	SDEBUG("   count: %d\n", count);
	if (arg == ATOM_UNIT_MICROSEC)
		udelay(count);
	else if (!drm_can_sleep())
		mdelay(count);
	else
		msleep(count);
}

static void atom_op_div(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src;
	SDEBUG("   src1: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, NULL, 1);
	SDEBUG("   src2: ");
	src = atom_get_src(ctx, attr, ptr);
	if (src != 0) {
		ctx->ctx->divmul[0] = dst / src;
		ctx->ctx->divmul[1] = dst % src;
	} else {
		ctx->ctx->divmul[0] = 0;
		ctx->ctx->divmul[1] = 0;
	}
}

static void atom_op_div32(atom_exec_context *ctx, int *ptr, int arg)
{
	uint64_t val64;
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src;
	SDEBUG("   src1: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, NULL, 1);
	SDEBUG("   src2: ");
	src = atom_get_src(ctx, attr, ptr);
	if (src != 0) {
		val64 = dst;
		val64 |= ((uint64_t)ctx->ctx->divmul[1]) << 32;
		do_div(val64, src);
		ctx->ctx->divmul[0] = lower_32_bits(val64);
		ctx->ctx->divmul[1] = upper_32_bits(val64);
	} else {
		ctx->ctx->divmul[0] = 0;
		ctx->ctx->divmul[1] = 0;
	}
}

static void atom_op_eot(atom_exec_context *ctx, int *ptr, int arg)
{
	/* functionally, a nop */
}

static void atom_op_jump(atom_exec_context *ctx, int *ptr, int arg)
{
	int execute = 0, target = U16(*ptr);
	unsigned long cjiffies;

	(*ptr) += 2;
	switch (arg) {
	case ATOM_COND_ABOVE:
		execute = ctx->ctx->cs_above;
		break;
	case ATOM_COND_ABOVEOREQUAL:
		execute = ctx->ctx->cs_above || ctx->ctx->cs_equal;
		break;
	case ATOM_COND_ALWAYS:
		execute = 1;
		break;
	case ATOM_COND_BELOW:
		execute = !(ctx->ctx->cs_above || ctx->ctx->cs_equal);
		break;
	case ATOM_COND_BELOWOREQUAL:
		execute = !ctx->ctx->cs_above;
		break;
	case ATOM_COND_EQUAL:
		execute = ctx->ctx->cs_equal;
		break;
	case ATOM_COND_NOTEQUAL:
		execute = !ctx->ctx->cs_equal;
		break;
	}
	if (arg != ATOM_COND_ALWAYS)
		SDEBUG("   taken: %s\n", execute ? "yes" : "no");
	SDEBUG("   target: 0x%04X\n", target);
	if (execute) {
		if (ctx->last_jump == (ctx->start + target)) {
			cjiffies = jiffies;
			if (time_after(cjiffies, ctx->last_jump_jiffies)) {
				cjiffies -= ctx->last_jump_jiffies;
				if ((jiffies_to_msecs(cjiffies) > ATOM_CMD_TIMEOUT_SEC*1000)) {
					DRM_ERROR("atombios stuck in loop for more than %dsecs aborting\n",
						  ATOM_CMD_TIMEOUT_SEC);
					ctx->abort = true;
				}
			} else {
				/* jiffies wrap around we will just wait a little longer */
				ctx->last_jump_jiffies = jiffies;
			}
		} else {
			ctx->last_jump = ctx->start + target;
			ctx->last_jump_jiffies = jiffies;
		}
		*ptr = ctx->start + target;
	}
}

static void atom_op_mask(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, mask, src, saved;
	int dptr = *ptr;
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	mask = atom_get_src_direct(ctx, ((attr >> 3) & 7), ptr);
	SDEBUG("   mask: 0x%08x", mask);
	SDEBUG("   src: ");
	src = atom_get_src(ctx, attr, ptr);
	dst &= mask;
	dst |= src;
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_move(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t src, saved;
	int dptr = *ptr;
	if (((attr >> 3) & 7) != ATOM_SRC_DWORD)
		atom_get_dst(ctx, arg, attr, ptr, &saved, 0);
	else {
		atom_skip_dst(ctx, arg, attr, ptr);
		saved = 0xCDCDCDCD;
	}
	SDEBUG("   src: ");
	src = atom_get_src(ctx, attr, ptr);
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, src, saved);
}

static void atom_op_mul(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src;
	SDEBUG("   src1: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, NULL, 1);
	SDEBUG("   src2: ");
	src = atom_get_src(ctx, attr, ptr);
	ctx->ctx->divmul[0] = dst * src;
}

static void atom_op_mul32(atom_exec_context *ctx, int *ptr, int arg)
{
	uint64_t val64;
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src;
	SDEBUG("   src1: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, NULL, 1);
	SDEBUG("   src2: ");
	src = atom_get_src(ctx, attr, ptr);
	val64 = (uint64_t)dst * (uint64_t)src;
	ctx->ctx->divmul[0] = lower_32_bits(val64);
	ctx->ctx->divmul[1] = upper_32_bits(val64);
}

static void atom_op_nop(atom_exec_context *ctx, int *ptr, int arg)
{
	/* nothing */
}

static void atom_op_or(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src, saved;
	int dptr = *ptr;
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	SDEBUG("   src: ");
	src = atom_get_src(ctx, attr, ptr);
	dst |= src;
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_postcard(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t val = U8((*ptr)++);
	SDEBUG("POST card output: 0x%02X\n", val);
}

static void atom_op_repeat(atom_exec_context *ctx, int *ptr, int arg)
{
	pr_info("unimplemented!\n");
}

static void atom_op_restorereg(atom_exec_context *ctx, int *ptr, int arg)
{
	pr_info("unimplemented!\n");
}

static void atom_op_savereg(atom_exec_context *ctx, int *ptr, int arg)
{
	pr_info("unimplemented!\n");
}

static void atom_op_setdatablock(atom_exec_context *ctx, int *ptr, int arg)
{
	int idx = U8(*ptr);
	(*ptr)++;
	SDEBUG("   block: %d\n", idx);
	if (!idx)
		ctx->ctx->data_block = 0;
	else if (idx == 255)
		ctx->ctx->data_block = ctx->start;
	else
		ctx->ctx->data_block = U16(ctx->ctx->data_table + 4 + 2 * idx);
	SDEBUG("   base: 0x%04X\n", ctx->ctx->data_block);
}

static void atom_op_setfbbase(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	SDEBUG("   fb_base: ");
	ctx->ctx->fb_base = atom_get_src(ctx, attr, ptr);
}

static void atom_op_setport(atom_exec_context *ctx, int *ptr, int arg)
{
	int port;
	switch (arg) {
	case ATOM_PORT_ATI:
		port = U16(*ptr);
		if (port < ATOM_IO_NAMES_CNT)
			SDEBUG("   port: %d (%s)\n", port, atom_io_names[port]);
		else
			SDEBUG("   port: %d\n", port);
		if (!port)
			ctx->ctx->io_mode = ATOM_IO_MM;
		else
			ctx->ctx->io_mode = ATOM_IO_IIO | port;
		(*ptr) += 2;
		break;
	case ATOM_PORT_PCI:
		ctx->ctx->io_mode = ATOM_IO_PCI;
		(*ptr)++;
		break;
	case ATOM_PORT_SYSIO:
		ctx->ctx->io_mode = ATOM_IO_SYSIO;
		(*ptr)++;
		break;
	}
}

static void atom_op_setregblock(atom_exec_context *ctx, int *ptr, int arg)
{
	ctx->ctx->reg_block = U16(*ptr);
	(*ptr) += 2;
	SDEBUG("   base: 0x%04X\n", ctx->ctx->reg_block);
}

static void atom_op_shift_left(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++), shift;
	uint32_t saved, dst;
	int dptr = *ptr;
	attr &= 0x38;
	attr |= atom_def_dst[attr >> 3] << 6;
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	shift = atom_get_src_direct(ctx, ATOM_SRC_BYTE0, ptr);
	SDEBUG("   shift: %d\n", shift);
	dst <<= shift;
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_shift_right(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++), shift;
	uint32_t saved, dst;
	int dptr = *ptr;
	attr &= 0x38;
	attr |= atom_def_dst[attr >> 3] << 6;
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	shift = atom_get_src_direct(ctx, ATOM_SRC_BYTE0, ptr);
	SDEBUG("   shift: %d\n", shift);
	dst >>= shift;
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_shl(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++), shift;
	uint32_t saved, dst;
	int dptr = *ptr;
	uint32_t dst_align = atom_dst_to_src[(attr >> 3) & 7][(attr >> 6) & 3];
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	/* op needs to full dst value */
	dst = saved;
	shift = atom_get_src(ctx, attr, ptr);
	SDEBUG("   shift: %d\n", shift);
	dst <<= shift;
	dst &= atom_arg_mask[dst_align];
	dst >>= atom_arg_shift[dst_align];
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_shr(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++), shift;
	uint32_t saved, dst;
	int dptr = *ptr;
	uint32_t dst_align = atom_dst_to_src[(attr >> 3) & 7][(attr >> 6) & 3];
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	/* op needs to full dst value */
	dst = saved;
	shift = atom_get_src(ctx, attr, ptr);
	SDEBUG("   shift: %d\n", shift);
	dst >>= shift;
	dst &= atom_arg_mask[dst_align];
	dst >>= atom_arg_shift[dst_align];
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_sub(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src, saved;
	int dptr = *ptr;
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	SDEBUG("   src: ");
	src = atom_get_src(ctx, attr, ptr);
	dst -= src;
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_switch(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t src, val, target;
	SDEBUG("   switch: ");
	src = atom_get_src(ctx, attr, ptr);
	while (U16(*ptr) != ATOM_CASE_END)
		if (U8(*ptr) == ATOM_CASE_MAGIC) {
			(*ptr)++;
			SDEBUG("   case: ");
			val =
			    atom_get_src(ctx, (attr & 0x38) | ATOM_ARG_IMM,
					 ptr);
			target = U16(*ptr);
			if (val == src) {
				SDEBUG("   target: %04X\n", target);
				*ptr = ctx->start + target;
				return;
			}
			(*ptr) += 2;
		} else {
			pr_info("Bad case\n");
			return;
		}
	(*ptr) += 2;
}

static void atom_op_test(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src;
	SDEBUG("   src1: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, NULL, 1);
	SDEBUG("   src2: ");
	src = atom_get_src(ctx, attr, ptr);
	ctx->ctx->cs_equal = ((dst & src) == 0);
	SDEBUG("   result: %s\n", ctx->ctx->cs_equal ? "EQ" : "NE");
}

static void atom_op_xor(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t attr = U8((*ptr)++);
	uint32_t dst, src, saved;
	int dptr = *ptr;
	SDEBUG("   dst: ");
	dst = atom_get_dst(ctx, arg, attr, ptr, &saved, 1);
	SDEBUG("   src: ");
	src = atom_get_src(ctx, attr, ptr);
	dst ^= src;
	SDEBUG("   dst: ");
	atom_put_dst(ctx, arg, attr, &dptr, dst, saved);
}

static void atom_op_debug(atom_exec_context *ctx, int *ptr, int arg)
{
	uint8_t val = U8((*ptr)++);
	SDEBUG("DEBUG output: 0x%02X\n", val);
}

static void atom_op_processds(atom_exec_context *ctx, int *ptr, int arg)
{
	uint16_t val = U16(*ptr);
	(*ptr) += val + 2;
	SDEBUG("PROCESSDS output: 0x%02X\n", val);
}

static struct {
	void (*func) (atom_exec_context *, int *, int);
	int arg;
} opcode_table[ATOM_OP_CNT] = {
	{
	NULL, 0}, {
	atom_op_move, ATOM_ARG_REG}, {
	atom_op_move, ATOM_ARG_PS}, {
	atom_op_move, ATOM_ARG_WS}, {
	atom_op_move, ATOM_ARG_FB}, {
	atom_op_move, ATOM_ARG_PLL}, {
	atom_op_move, ATOM_ARG_MC}, {
	atom_op_and, ATOM_ARG_REG}, {
	atom_op_and, ATOM_ARG_PS}, {
	atom_op_and, ATOM_ARG_WS}, {
	atom_op_and, ATOM_ARG_FB}, {
	atom_op_and, ATOM_ARG_PLL}, {
	atom_op_and, ATOM_ARG_MC}, {
	atom_op_or, ATOM_ARG_REG}, {
	atom_op_or, ATOM_ARG_PS}, {
	atom_op_or, ATOM_ARG_WS}, {
	atom_op_or, ATOM_ARG_FB}, {
	atom_op_or, ATOM_ARG_PLL}, {
	atom_op_or, ATOM_ARG_MC}, {
	atom_op_shift_left, ATOM_ARG_REG}, {
	atom_op_shift_left, ATOM_ARG_PS}, {
	atom_op_shift_left, ATOM_ARG_WS}, {
	atom_op_shift_left, ATOM_ARG_FB}, {
	atom_op_shift_left, ATOM_ARG_PLL}, {
	atom_op_shift_left, ATOM_ARG_MC}, {
	atom_op_shift_right, ATOM_ARG_REG}, {
	atom_op_shift_right, ATOM_ARG_PS}, {
	atom_op_shift_right, ATOM_ARG_WS}, {
	atom_op_shift_right, ATOM_ARG_FB}, {
	atom_op_shift_right, ATOM_ARG_PLL}, {
	atom_op_shift_right, ATOM_ARG_MC}, {
	atom_op_mul, ATOM_ARG_REG}, {
	atom_op_mul, ATOM_ARG_PS}, {
	atom_op_mul, ATOM_ARG_WS}, {
	atom_op_mul, ATOM_ARG_FB}, {
	atom_op_mul, ATOM_ARG_PLL}, {
	atom_op_mul, ATOM_ARG_MC}, {
	atom_op_div, ATOM_ARG_REG}, {
	atom_op_div, ATOM_ARG_PS}, {
	atom_op_div, ATOM_ARG_WS}, {
	atom_op_div, ATOM_ARG_FB}, {
	atom_op_div, ATOM_ARG_PLL}, {
	atom_op_div, ATOM_ARG_MC}, {
	atom_op_add, ATOM_ARG_REG}, {
	atom_op_add, ATOM_ARG_PS}, {
	atom_op_add, ATOM_ARG_WS}, {
	atom_op_add, ATOM_ARG_FB}, {
	atom_op_add, ATOM_ARG_PLL}, {
	atom_op_add, ATOM_ARG_MC}, {
	atom_op_sub, ATOM_ARG_REG}, {
	atom_op_sub, ATOM_ARG_PS}, {
	atom_op_sub, ATOM_ARG_WS}, {
	atom_op_sub, ATOM_ARG_FB}, {
	atom_op_sub, ATOM_ARG_PLL}, {
	atom_op_sub, ATOM_ARG_MC}, {
	atom_op_setport, ATOM_PORT_ATI}, {
	atom_op_setport, ATOM_PORT_PCI}, {
	atom_op_setport, ATOM_PORT_SYSIO}, {
	atom_op_setregblock, 0}, {
	atom_op_setfbbase, 0}, {
	atom_op_compare, ATOM_ARG_REG}, {
	atom_op_compare, ATOM_ARG_PS}, {
	atom_op_compare, ATOM_ARG_WS}, {
	atom_op_compare, ATOM_ARG_FB}, {
	atom_op_compare, ATOM_ARG_PLL}, {
	atom_op_compare, ATOM_ARG_MC}, {
	atom_op_switch, 0}, {
	atom_op_jump, ATOM_COND_ALWAYS}, {
	atom_op_jump, ATOM_COND_EQUAL}, {
	atom_op_jump, ATOM_COND_BELOW}, {
	atom_op_jump, ATOM_COND_ABOVE}, {
	atom_op_jump, ATOM_COND_BELOWOREQUAL}, {
	atom_op_jump, ATOM_COND_ABOVEOREQUAL}, {
	atom_op_jump, ATOM_COND_NOTEQUAL}, {
	atom_op_test, ATOM_ARG_REG}, {
	atom_op_test, ATOM_ARG_PS}, {
	atom_op_test, ATOM_ARG_WS}, {
	atom_op_test, ATOM_ARG_FB}, {
	atom_op_test, ATOM_ARG_PLL}, {
	atom_op_test, ATOM_ARG_MC}, {
	atom_op_delay, ATOM_UNIT_MILLISEC}, {
	atom_op_delay, ATOM_UNIT_MICROSEC}, {
	atom_op_calltable, 0}, {
	atom_op_repeat, 0}, {
	atom_op_clear, ATOM_ARG_REG}, {
	atom_op_clear, ATOM_ARG_PS}, {
	atom_op_clear, ATOM_ARG_WS}, {
	atom_op_clear, ATOM_ARG_FB}, {
	atom_op_clear, ATOM_ARG_PLL}, {
	atom_op_clear, ATOM_ARG_MC}, {
	atom_op_nop, 0}, {
	atom_op_eot, 0}, {
	atom_op_mask, ATOM_ARG_REG}, {
	atom_op_mask, ATOM_ARG_PS}, {
	atom_op_mask, ATOM_ARG_WS}, {
	atom_op_mask, ATOM_ARG_FB}, {
	atom_op_mask, ATOM_ARG_PLL}, {
	atom_op_mask, ATOM_ARG_MC}, {
	atom_op_postcard, 0}, {
	atom_op_beep, 0}, {
	atom_op_savereg, 0}, {
	atom_op_restorereg, 0}, {
	atom_op_setdatablock, 0}, {
	atom_op_xor, ATOM_ARG_REG}, {
	atom_op_xor, ATOM_ARG_PS}, {
	atom_op_xor, ATOM_ARG_WS}, {
	atom_op_xor, ATOM_ARG_FB}, {
	atom_op_xor, ATOM_ARG_PLL}, {
	atom_op_xor, ATOM_ARG_MC}, {
	atom_op_shl, ATOM_ARG_REG}, {
	atom_op_shl, ATOM_ARG_PS}, {
	atom_op_shl, ATOM_ARG_WS}, {
	atom_op_shl, ATOM_ARG_FB}, {
	atom_op_shl, ATOM_ARG_PLL}, {
	atom_op_shl, ATOM_ARG_MC}, {
	atom_op_shr, ATOM_ARG_REG}, {
	atom_op_shr, ATOM_ARG_PS}, {
	atom_op_shr, ATOM_ARG_WS}, {
	atom_op_shr, ATOM_ARG_FB}, {
	atom_op_shr, ATOM_ARG_PLL}, {
	atom_op_shr, ATOM_ARG_MC}, {
	atom_op_debug, 0}, {
	atom_op_processds, 0}, {
	atom_op_mul32, ATOM_ARG_PS}, {
	atom_op_mul32, ATOM_ARG_WS}, {
	atom_op_div32, ATOM_ARG_PS}, {
	atom_op_div32, ATOM_ARG_WS},
};

static int amdgpu_atom_execute_table_locked(struct atom_context *ctx, int index, uint32_t * params)
{
	int base = CU16(ctx->cmd_table + 4 + 2 * index);
	int len, ws, ps, ptr;
	unsigned char op;
	atom_exec_context ectx;
	int ret = 0;

	if (!base)
		return -EINVAL;

	len = CU16(base + ATOM_CT_SIZE_PTR);
	ws = CU8(base + ATOM_CT_WS_PTR);
	ps = CU8(base + ATOM_CT_PS_PTR) & ATOM_CT_PS_MASK;
	ptr = base + ATOM_CT_CODE_PTR;

	SDEBUG(">> execute %04X (len %d, WS %d, PS %d)\n", base, len, ws, ps);

	ectx.ctx = ctx;
	ectx.ps_shift = ps / 4;
	ectx.start = base;
	ectx.ps = params;
	ectx.abort = false;
	ectx.last_jump = 0;
	if (ws)
		ectx.ws = kcalloc(4, ws, GFP_KERNEL);
	else
		ectx.ws = NULL;

	debug_depth++;
	while (1) {
		op = CU8(ptr++);
		if (op < ATOM_OP_NAMES_CNT)
			SDEBUG("%s @ 0x%04X\n", atom_op_names[op], ptr - 1);
		else
			SDEBUG("[%d] @ 0x%04X\n", op, ptr - 1);
		if (ectx.abort) {
			DRM_ERROR("atombios stuck executing %04X (len %d, WS %d, PS %d) @ 0x%04X\n",
				base, len, ws, ps, ptr - 1);
			ret = -EINVAL;
			goto free;
		}

		if (op < ATOM_OP_CNT && op > 0)
			opcode_table[op].func(&ectx, &ptr,
					      opcode_table[op].arg);
		else
			break;

		if (op == ATOM_OP_EOT)
			break;
	}
	debug_depth--;
	SDEBUG("<<\n");

free:
	if (ws)
		kfree(ectx.ws);
	return ret;
}

int amdgpu_atom_execute_table(struct atom_context *ctx, int index, uint32_t * params)
{
	int r;

	mutex_lock(&ctx->mutex);
	/* reset data block */
	ctx->data_block = 0;
	/* reset reg block */
	ctx->reg_block = 0;
	/* reset fb window */
	ctx->fb_base = 0;
	/* reset io mode */
	ctx->io_mode = ATOM_IO_MM;
	/* reset divmul */
	ctx->divmul[0] = 0;
	ctx->divmul[1] = 0;
	r = amdgpu_atom_execute_table_locked(ctx, index, params);
	mutex_unlock(&ctx->mutex);
	return r;
}

static int atom_iio_len[] = { 1, 2, 3, 3, 3, 3, 4, 4, 4, 3 };

static void atom_index_iio(struct atom_context *ctx, int base)
{
	ctx->iio = kzalloc(2 * 256, GFP_KERNEL);
	if (!ctx->iio)
		return;
	while (CU8(base) == ATOM_IIO_START) {
		ctx->iio[CU8(base + 1)] = base + 2;
		base += 2;
		while (CU8(base) != ATOM_IIO_END)
			base += atom_iio_len[CU8(base)];
		base += 3;
	}
}

struct atom_context *amdgpu_atom_parse(struct card_info *card, void *bios)
{
	int base;
	struct atom_context *ctx =
	    kzalloc(sizeof(struct atom_context), GFP_KERNEL);
	char *str;
	u16 idx;

	if (!ctx)
		return NULL;

	ctx->card = card;
	ctx->bios = bios;

	if (CU16(0) != ATOM_BIOS_MAGIC) {
		pr_info("Invalid BIOS magic\n");
		kfree(ctx);
		return NULL;
	}
	if (strncmp
	    (CSTR(ATOM_ATI_MAGIC_PTR), ATOM_ATI_MAGIC,
	     strlen(ATOM_ATI_MAGIC))) {
		pr_info("Invalid ATI magic\n");
		kfree(ctx);
		return NULL;
	}

	base = CU16(ATOM_ROM_TABLE_PTR);
	if (strncmp
	    (CSTR(base + ATOM_ROM_MAGIC_PTR), ATOM_ROM_MAGIC,
	     strlen(ATOM_ROM_MAGIC))) {
		pr_info("Invalid ATOM magic\n");
		kfree(ctx);
		return NULL;
	}

	ctx->cmd_table = CU16(base + ATOM_ROM_CMD_PTR);
	ctx->data_table = CU16(base + ATOM_ROM_DATA_PTR);
	atom_index_iio(ctx, CU16(ctx->data_table + ATOM_DATA_IIO_PTR) + 4);
	if (!ctx->iio) {
		amdgpu_atom_destroy(ctx);
		return NULL;
	}

	idx = CU16(ATOM_ROM_PART_NUMBER_PTR);
	if (idx == 0)
		idx = 0x80;

	str = CSTR(idx);
	if (*str != '\0') {
		pr_info("ATOM BIOS: %s\n", str);
		strlcpy(ctx->vbios_version, str, sizeof(ctx->vbios_version));
	}


	return ctx;
}

int amdgpu_atom_asic_init(struct atom_context *ctx)
{
	int hwi = CU16(ctx->data_table + ATOM_DATA_FWI_PTR);
	uint32_t ps[16];
	int ret;

	memset(ps, 0, 64);

	ps[0] = cpu_to_le32(CU32(hwi + ATOM_FWI_DEFSCLK_PTR));
	ps[1] = cpu_to_le32(CU32(hwi + ATOM_FWI_DEFMCLK_PTR));
	if (!ps[0] || !ps[1])
		return 1;

	if (!CU16(ctx->cmd_table + 4 + 2 * ATOM_CMD_INIT))
		return 1;
	ret = amdgpu_atom_execute_table(ctx, ATOM_CMD_INIT, ps);
	if (ret)
		return ret;

	memset(ps, 0, 64);

	return ret;
}

void amdgpu_atom_destroy(struct atom_context *ctx)
{
	kfree(ctx->iio);
	kfree(ctx);
}

bool amdgpu_atom_parse_data_header(struct atom_context *ctx, int index,
			    uint16_t * size, uint8_t * frev, uint8_t * crev,
			    uint16_t * data_start)
{
	int offset = index * 2 + 4;
	int idx = CU16(ctx->data_table + offset);
	u16 *mdt = (u16 *)(ctx->bios + ctx->data_table + 4);

	if (!mdt[index])
		return false;

	if (size)
		*size = CU16(idx);
	if (frev)
		*frev = CU8(idx + 2);
	if (crev)
		*crev = CU8(idx + 3);
	*data_start = idx;
	return true;
}

bool amdgpu_atom_parse_cmd_header(struct atom_context *ctx, int index, uint8_t * frev,
			   uint8_t * crev)
{
	int offset = index * 2 + 4;
	int idx = CU16(ctx->cmd_table + offset);
	u16 *mct = (u16 *)(ctx->bios + ctx->cmd_table + 4);

	if (!mct[index])
		return false;

	if (frev)
		*frev = CU8(idx + 2);
	if (crev)
		*crev = CU8(idx + 3);
	return true;
}