xref: /linux/drivers/gpu/drm/nouveau/nvkm/engine/gr/gk104.c (revision 975ef7ff81bb000af6e6c8e63e81f89f3468dcf7)
1 /*
2  * Copyright 2013 Red Hat Inc.
3  *
4  * Permission is hereby granted, free of charge, to any person obtaining a
5  * copy of this software and associated documentation files (the "Software"),
6  * to deal in the Software without restriction, including without limitation
7  * the rights to use, copy, modify, merge, publish, distribute, sublicense,
8  * and/or sell copies of the Software, and to permit persons to whom the
9  * Software is furnished to do so, subject to the following conditions:
10  *
11  * The above copyright notice and this permission notice shall be included in
12  * all copies or substantial portions of the Software.
13  *
14  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
15  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
16  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
17  * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
18  * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
19  * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
20  * OTHER DEALINGS IN THE SOFTWARE.
21  *
22  * Authors: Ben Skeggs <bskeggs@redhat.com>
23  */
24 #include "gf100.h"
25 #include "gk104.h"
26 #include "ctxgf100.h"
27 
28 #include <nvif/class.h>
29 
30 /*******************************************************************************
31  * PGRAPH register lists
32  ******************************************************************************/
33 
34 const struct gf100_gr_init
35 gk104_gr_init_main_0[] = {
36 	{ 0x400080,   1, 0x04, 0x003083c2 },
37 	{ 0x400088,   1, 0x04, 0x0001ffe7 },
38 	{ 0x40008c,   1, 0x04, 0x00000000 },
39 	{ 0x400090,   1, 0x04, 0x00000030 },
40 	{ 0x40013c,   1, 0x04, 0x003901f7 },
41 	{ 0x400140,   1, 0x04, 0x00000100 },
42 	{ 0x400144,   1, 0x04, 0x00000000 },
43 	{ 0x400148,   1, 0x04, 0x00000110 },
44 	{ 0x400138,   1, 0x04, 0x00000000 },
45 	{ 0x400130,   2, 0x04, 0x00000000 },
46 	{ 0x400124,   1, 0x04, 0x00000002 },
47 	{}
48 };
49 
50 static const struct gf100_gr_init
51 gk104_gr_init_ds_0[] = {
52 	{ 0x405844,   1, 0x04, 0x00ffffff },
53 	{ 0x405850,   1, 0x04, 0x00000000 },
54 	{ 0x405900,   1, 0x04, 0x0000ff34 },
55 	{ 0x405908,   1, 0x04, 0x00000000 },
56 	{ 0x405928,   2, 0x04, 0x00000000 },
57 	{}
58 };
59 
60 static const struct gf100_gr_init
61 gk104_gr_init_sked_0[] = {
62 	{ 0x407010,   1, 0x04, 0x00000000 },
63 	{}
64 };
65 
66 static const struct gf100_gr_init
67 gk104_gr_init_cwd_0[] = {
68 	{ 0x405b50,   1, 0x04, 0x00000000 },
69 	{}
70 };
71 
72 static const struct gf100_gr_init
73 gk104_gr_init_gpc_unk_1[] = {
74 	{ 0x418d00,   1, 0x04, 0x00000000 },
75 	{ 0x418d28,   2, 0x04, 0x00000000 },
76 	{ 0x418f00,   1, 0x04, 0x00000000 },
77 	{ 0x418f08,   1, 0x04, 0x00000000 },
78 	{ 0x418f20,   2, 0x04, 0x00000000 },
79 	{ 0x418e00,   1, 0x04, 0x00000060 },
80 	{ 0x418e08,   1, 0x04, 0x00000000 },
81 	{ 0x418e1c,   2, 0x04, 0x00000000 },
82 	{}
83 };
84 
85 const struct gf100_gr_init
86 gk104_gr_init_gpc_unk_2[] = {
87 	{ 0x418884,   1, 0x04, 0x00000000 },
88 	{}
89 };
90 
91 const struct gf100_gr_init
92 gk104_gr_init_tpccs_0[] = {
93 	{ 0x419d0c,   1, 0x04, 0x00000000 },
94 	{ 0x419d10,   1, 0x04, 0x00000014 },
95 	{}
96 };
97 
98 const struct gf100_gr_init
99 gk104_gr_init_pe_0[] = {
100 	{ 0x41980c,   1, 0x04, 0x00000010 },
101 	{ 0x419844,   1, 0x04, 0x00000000 },
102 	{ 0x419850,   1, 0x04, 0x00000004 },
103 	{ 0x419854,   2, 0x04, 0x00000000 },
104 	{}
105 };
106 
107 static const struct gf100_gr_init
108 gk104_gr_init_l1c_0[] = {
109 	{ 0x419c98,   1, 0x04, 0x00000000 },
110 	{ 0x419ca8,   1, 0x04, 0x00000000 },
111 	{ 0x419cb0,   1, 0x04, 0x01000000 },
112 	{ 0x419cb4,   1, 0x04, 0x00000000 },
113 	{ 0x419cb8,   1, 0x04, 0x00b08bea },
114 	{ 0x419c84,   1, 0x04, 0x00010384 },
115 	{ 0x419cbc,   1, 0x04, 0x28137646 },
116 	{ 0x419cc0,   2, 0x04, 0x00000000 },
117 	{ 0x419c80,   1, 0x04, 0x00020232 },
118 	{}
119 };
120 
121 static const struct gf100_gr_init
122 gk104_gr_init_sm_0[] = {
123 	{ 0x419e00,   1, 0x04, 0x00000000 },
124 	{ 0x419ea0,   1, 0x04, 0x00000000 },
125 	{ 0x419ee4,   1, 0x04, 0x00000000 },
126 	{ 0x419ea4,   1, 0x04, 0x00000100 },
127 	{ 0x419ea8,   1, 0x04, 0x00000000 },
128 	{ 0x419eb4,   4, 0x04, 0x00000000 },
129 	{ 0x419edc,   1, 0x04, 0x00000000 },
130 	{ 0x419f00,   1, 0x04, 0x00000000 },
131 	{ 0x419f74,   1, 0x04, 0x00000555 },
132 	{}
133 };
134 
135 const struct gf100_gr_init
136 gk104_gr_init_be_0[] = {
137 	{ 0x40880c,   1, 0x04, 0x00000000 },
138 	{ 0x408850,   1, 0x04, 0x00000004 },
139 	{ 0x408910,   9, 0x04, 0x00000000 },
140 	{ 0x408950,   1, 0x04, 0x00000000 },
141 	{ 0x408954,   1, 0x04, 0x0000ffff },
142 	{ 0x408958,   1, 0x04, 0x00000034 },
143 	{ 0x408984,   1, 0x04, 0x00000000 },
144 	{ 0x408988,   1, 0x04, 0x08040201 },
145 	{ 0x40898c,   1, 0x04, 0x80402010 },
146 	{}
147 };
148 
149 const struct gf100_gr_pack
150 gk104_gr_pack_mmio[] = {
151 	{ gk104_gr_init_main_0 },
152 	{ gf100_gr_init_fe_0 },
153 	{ gf100_gr_init_pri_0 },
154 	{ gf100_gr_init_rstr2d_0 },
155 	{ gf119_gr_init_pd_0 },
156 	{ gk104_gr_init_ds_0 },
157 	{ gf100_gr_init_scc_0 },
158 	{ gk104_gr_init_sked_0 },
159 	{ gk104_gr_init_cwd_0 },
160 	{ gf119_gr_init_prop_0 },
161 	{ gf108_gr_init_gpc_unk_0 },
162 	{ gf100_gr_init_setup_0 },
163 	{ gf100_gr_init_crstr_0 },
164 	{ gf108_gr_init_setup_1 },
165 	{ gf100_gr_init_zcull_0 },
166 	{ gf119_gr_init_gpm_0 },
167 	{ gk104_gr_init_gpc_unk_1 },
168 	{ gf100_gr_init_gcc_0 },
169 	{ gk104_gr_init_gpc_unk_2 },
170 	{ gk104_gr_init_tpccs_0 },
171 	{ gf119_gr_init_tex_0 },
172 	{ gk104_gr_init_pe_0 },
173 	{ gk104_gr_init_l1c_0 },
174 	{ gf100_gr_init_mpc_0 },
175 	{ gk104_gr_init_sm_0 },
176 	{ gf117_gr_init_pes_0 },
177 	{ gf117_gr_init_wwdx_0 },
178 	{ gf117_gr_init_cbm_0 },
179 	{ gk104_gr_init_be_0 },
180 	{ gf100_gr_init_fe_1 },
181 	{}
182 };
183 
184 const struct nvkm_therm_clkgate_init
185 gk104_clkgate_blcg_init_main_0[] = {
186 	{ 0x4041f0, 1, 0x00004046 },
187 	{ 0x409890, 1, 0x00000045 },
188 	{ 0x4098b0, 1, 0x0000007f },
189 	{}
190 };
191 
192 const struct nvkm_therm_clkgate_init
193 gk104_clkgate_blcg_init_rstr2d_0[] = {
194 	{ 0x4078c0, 1, 0x00000042 },
195 	{}
196 };
197 
198 const struct nvkm_therm_clkgate_init
199 gk104_clkgate_blcg_init_unk_0[] = {
200 	{ 0x406000, 1, 0x00004044 },
201 	{ 0x405860, 1, 0x00004042 },
202 	{ 0x40590c, 1, 0x00004042 },
203 	{}
204 };
205 
206 const struct nvkm_therm_clkgate_init
207 gk104_clkgate_blcg_init_gcc_0[] = {
208 	{ 0x408040, 1, 0x00004044 },
209 	{}
210 };
211 
212 const struct nvkm_therm_clkgate_init
213 gk104_clkgate_blcg_init_sked_0[] = {
214 	{ 0x407000, 1, 0x00004044 },
215 	{}
216 };
217 
218 const struct nvkm_therm_clkgate_init
219 gk104_clkgate_blcg_init_unk_1[] = {
220 	{ 0x405bf0, 1, 0x00004044 },
221 	{}
222 };
223 
224 const struct nvkm_therm_clkgate_init
225 gk104_clkgate_blcg_init_gpc_ctxctl_0[] = {
226 	{ 0x41a890, 1, 0x00000042 },
227 	{ 0x41a8b0, 1, 0x0000007f },
228 	{}
229 };
230 
231 const struct nvkm_therm_clkgate_init
232 gk104_clkgate_blcg_init_gpc_unk_0[] = {
233 	{ 0x418500, 1, 0x00004042 },
234 	{ 0x418608, 1, 0x00004042 },
235 	{ 0x418688, 1, 0x00004042 },
236 	{ 0x418718, 1, 0x00000042 },
237 	{}
238 };
239 
240 const struct nvkm_therm_clkgate_init
241 gk104_clkgate_blcg_init_gpc_esetup_0[] = {
242 	{ 0x418828, 1, 0x00000044 },
243 	{}
244 };
245 
246 const struct nvkm_therm_clkgate_init
247 gk104_clkgate_blcg_init_gpc_tpbus_0[] = {
248 	{ 0x418bbc, 1, 0x00004042 },
249 	{}
250 };
251 
252 const struct nvkm_therm_clkgate_init
253 gk104_clkgate_blcg_init_gpc_zcull_0[] = {
254 	{ 0x418970, 1, 0x00004042 },
255 	{}
256 };
257 
258 const struct nvkm_therm_clkgate_init
259 gk104_clkgate_blcg_init_gpc_tpconf_0[] = {
260 	{ 0x418c70, 1, 0x00004042 },
261 	{}
262 };
263 
264 const struct nvkm_therm_clkgate_init
265 gk104_clkgate_blcg_init_gpc_unk_1[] = {
266 	{ 0x418cf0, 1, 0x00004042 },
267 	{ 0x418d70, 1, 0x00004042 },
268 	{ 0x418f0c, 1, 0x00004042 },
269 	{ 0x418e0c, 1, 0x00004042 },
270 	{}
271 };
272 
273 const struct nvkm_therm_clkgate_init
274 gk104_clkgate_blcg_init_gpc_gcc_0[] = {
275 	{ 0x419020, 1, 0x00004042 },
276 	{ 0x419038, 1, 0x00000042 },
277 	{}
278 };
279 
280 const struct nvkm_therm_clkgate_init
281 gk104_clkgate_blcg_init_gpc_ffb_0[] = {
282 	{ 0x418898, 1, 0x00000042 },
283 	{}
284 };
285 
286 const struct nvkm_therm_clkgate_init
287 gk104_clkgate_blcg_init_gpc_tex_0[] = {
288 	{ 0x419a40, 9, 0x00004042 },
289 	{ 0x419acc, 1, 0x00004047 },
290 	{}
291 };
292 
293 const struct nvkm_therm_clkgate_init
294 gk104_clkgate_blcg_init_gpc_poly_0[] = {
295 	{ 0x419868, 1, 0x00000042 },
296 	{}
297 };
298 
299 const struct nvkm_therm_clkgate_init
300 gk104_clkgate_blcg_init_gpc_l1c_0[] = {
301 	{ 0x419ccc, 3, 0x00000042 },
302 	{}
303 };
304 
305 const struct nvkm_therm_clkgate_init
306 gk104_clkgate_blcg_init_gpc_unk_2[] = {
307 	{ 0x419c70, 1, 0x00004045 },
308 	{}
309 };
310 
311 const struct nvkm_therm_clkgate_init
312 gk104_clkgate_blcg_init_gpc_mp_0[] = {
313 	{ 0x419fd0, 1, 0x00004043 },
314 	{ 0x419fd8, 1, 0x00004049 },
315 	{ 0x419fe0, 2, 0x00004042 },
316 	{ 0x419ff0, 1, 0x00004046 },
317 	{ 0x419ff8, 1, 0x00004042 },
318 	{}
319 };
320 
321 const struct nvkm_therm_clkgate_init
322 gk104_clkgate_blcg_init_gpc_ppc_0[] = {
323 	{ 0x41be28, 1, 0x00000042 },
324 	{ 0x41bfe8, 1, 0x00004042 },
325 	{ 0x41bed0, 1, 0x00004042 },
326 	{}
327 };
328 
329 const struct nvkm_therm_clkgate_init
330 gk104_clkgate_blcg_init_rop_zrop_0[] = {
331 	{ 0x408810, 2, 0x00004042 },
332 	{}
333 };
334 
335 const struct nvkm_therm_clkgate_init
336 gk104_clkgate_blcg_init_rop_0[] = {
337 	{ 0x408a80, 6, 0x00004042 },
338 	{}
339 };
340 
341 const struct nvkm_therm_clkgate_init
342 gk104_clkgate_blcg_init_rop_crop_0[] = {
343 	{ 0x4089a8, 1, 0x00004042 },
344 	{ 0x4089b0, 1, 0x00000042 },
345 	{ 0x4089b8, 1, 0x00004042 },
346 	{}
347 };
348 
349 const struct nvkm_therm_clkgate_init
350 gk104_clkgate_blcg_init_pxbar_0[] = {
351 	{ 0x13c820, 1, 0x0001007f },
352 	{ 0x13cbe0, 1, 0x00000042 },
353 	{}
354 };
355 
356 static const struct nvkm_therm_clkgate_pack
357 gk104_clkgate_pack[] = {
358 	{ gk104_clkgate_blcg_init_main_0 },
359 	{ gk104_clkgate_blcg_init_rstr2d_0 },
360 	{ gk104_clkgate_blcg_init_unk_0 },
361 	{ gk104_clkgate_blcg_init_gcc_0 },
362 	{ gk104_clkgate_blcg_init_sked_0 },
363 	{ gk104_clkgate_blcg_init_unk_1 },
364 	{ gk104_clkgate_blcg_init_gpc_ctxctl_0 },
365 	{ gk104_clkgate_blcg_init_gpc_unk_0 },
366 	{ gk104_clkgate_blcg_init_gpc_esetup_0 },
367 	{ gk104_clkgate_blcg_init_gpc_tpbus_0 },
368 	{ gk104_clkgate_blcg_init_gpc_zcull_0 },
369 	{ gk104_clkgate_blcg_init_gpc_tpconf_0 },
370 	{ gk104_clkgate_blcg_init_gpc_unk_1 },
371 	{ gk104_clkgate_blcg_init_gpc_gcc_0 },
372 	{ gk104_clkgate_blcg_init_gpc_ffb_0 },
373 	{ gk104_clkgate_blcg_init_gpc_tex_0 },
374 	{ gk104_clkgate_blcg_init_gpc_poly_0 },
375 	{ gk104_clkgate_blcg_init_gpc_l1c_0 },
376 	{ gk104_clkgate_blcg_init_gpc_unk_2 },
377 	{ gk104_clkgate_blcg_init_gpc_mp_0 },
378 	{ gk104_clkgate_blcg_init_gpc_ppc_0 },
379 	{ gk104_clkgate_blcg_init_rop_zrop_0 },
380 	{ gk104_clkgate_blcg_init_rop_0 },
381 	{ gk104_clkgate_blcg_init_rop_crop_0 },
382 	{ gk104_clkgate_blcg_init_pxbar_0 },
383 	{}
384 };
385 
386 /*******************************************************************************
387  * PGRAPH engine/subdev functions
388  ******************************************************************************/
389 
390 void
391 gk104_gr_init_sked_hww_esr(struct gf100_gr *gr)
392 {
393 	nvkm_wr32(gr->base.engine.subdev.device, 0x407020, 0x40000000);
394 }
395 
396 static void
397 gk104_gr_init_fecs_exceptions(struct gf100_gr *gr)
398 {
399 	struct nvkm_device *device = gr->base.engine.subdev.device;
400 	nvkm_wr32(device, 0x409ffc, 0x00000000);
401 	nvkm_wr32(device, 0x409c14, 0x00003e3e);
402 	nvkm_wr32(device, 0x409c24, 0x000f0001);
403 }
404 
405 void
406 gk104_gr_init_rop_active_fbps(struct gf100_gr *gr)
407 {
408 	struct nvkm_device *device = gr->base.engine.subdev.device;
409 	const u32 fbp_count = nvkm_rd32(device, 0x120074);
410 	nvkm_mask(device, 0x408850, 0x0000000f, fbp_count); /* zrop */
411 	nvkm_mask(device, 0x408958, 0x0000000f, fbp_count); /* crop */
412 }
413 
414 void
415 gk104_gr_init_ppc_exceptions(struct gf100_gr *gr)
416 {
417 	struct nvkm_device *device = gr->base.engine.subdev.device;
418 	int gpc, ppc;
419 
420 	for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
421 		for (ppc = 0; ppc < gr->ppc_nr[gpc]; ppc++) {
422 			if (!(gr->ppc_mask[gpc] & (1 << ppc)))
423 				continue;
424 			nvkm_wr32(device, PPC_UNIT(gpc, ppc, 0x038), 0xc0000000);
425 		}
426 	}
427 }
428 
429 void
430 gk104_gr_init_vsc_stream_master(struct gf100_gr *gr)
431 {
432 	struct nvkm_device *device = gr->base.engine.subdev.device;
433 	nvkm_wr32(device, GPC_UNIT(0, 0x3018), 0x00000001);
434 }
435 
436 #include "fuc/hubgk104.fuc3.h"
437 
438 static struct gf100_gr_ucode
439 gk104_gr_fecs_ucode = {
440 	.code.data = gk104_grhub_code,
441 	.code.size = sizeof(gk104_grhub_code),
442 	.data.data = gk104_grhub_data,
443 	.data.size = sizeof(gk104_grhub_data),
444 };
445 
446 #include "fuc/gpcgk104.fuc3.h"
447 
448 static struct gf100_gr_ucode
449 gk104_gr_gpccs_ucode = {
450 	.code.data = gk104_grgpc_code,
451 	.code.size = sizeof(gk104_grgpc_code),
452 	.data.data = gk104_grgpc_data,
453 	.data.size = sizeof(gk104_grgpc_data),
454 };
455 
456 static const struct gf100_gr_func
457 gk104_gr = {
458 	.oneinit_tiles = gf100_gr_oneinit_tiles,
459 	.oneinit_sm_id = gf100_gr_oneinit_sm_id,
460 	.init = gf100_gr_init,
461 	.init_gpc_mmu = gf100_gr_init_gpc_mmu,
462 	.init_vsc_stream_master = gk104_gr_init_vsc_stream_master,
463 	.init_zcull = gf117_gr_init_zcull,
464 	.init_num_active_ltcs = gf100_gr_init_num_active_ltcs,
465 	.init_rop_active_fbps = gk104_gr_init_rop_active_fbps,
466 	.init_fecs_exceptions = gk104_gr_init_fecs_exceptions,
467 	.init_sked_hww_esr = gk104_gr_init_sked_hww_esr,
468 	.init_419cc0 = gf100_gr_init_419cc0,
469 	.init_419eb4 = gf100_gr_init_419eb4,
470 	.init_ppc_exceptions = gk104_gr_init_ppc_exceptions,
471 	.init_tex_hww_esr = gf100_gr_init_tex_hww_esr,
472 	.init_shader_exceptions = gf100_gr_init_shader_exceptions,
473 	.init_400054 = gf100_gr_init_400054,
474 	.trap_mp = gf100_gr_trap_mp,
475 	.mmio = gk104_gr_pack_mmio,
476 	.fecs.ucode = &gk104_gr_fecs_ucode,
477 	.gpccs.ucode = &gk104_gr_gpccs_ucode,
478 	.rops = gf100_gr_rops,
479 	.ppc_nr = 1,
480 	.grctx = &gk104_grctx,
481 	.clkgate_pack = gk104_clkgate_pack,
482 	.zbc = &gf100_gr_zbc,
483 	.sclass = {
484 		{ -1, -1, FERMI_TWOD_A },
485 		{ -1, -1, KEPLER_INLINE_TO_MEMORY_A },
486 		{ -1, -1, KEPLER_A, &gf100_fermi },
487 		{ -1, -1, KEPLER_COMPUTE_A },
488 		{}
489 	}
490 };
491 
492 int
493 gk104_gr_new(struct nvkm_device *device, int index, struct nvkm_gr **pgr)
494 {
495 	return gf100_gr_new_(&gk104_gr, device, index, pgr);
496 }
497