rjw | 1f88458 | 2022-01-06 17:20:42 +0800 | [diff] [blame^] | 1 | /* |
| 2 | * Tegra host1x Job |
| 3 | * |
| 4 | * Copyright (c) 2010-2015, NVIDIA Corporation. |
| 5 | * |
| 6 | * This program is free software; you can redistribute it and/or modify it |
| 7 | * under the terms and conditions of the GNU General Public License, |
| 8 | * version 2, as published by the Free Software Foundation. |
| 9 | * |
| 10 | * This program is distributed in the hope it will be useful, but WITHOUT |
| 11 | * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or |
| 12 | * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for |
| 13 | * more details. |
| 14 | * |
| 15 | * You should have received a copy of the GNU General Public License |
| 16 | * along with this program. If not, see <http://www.gnu.org/licenses/>. |
| 17 | */ |
| 18 | |
| 19 | #include <linux/dma-mapping.h> |
| 20 | #include <linux/err.h> |
| 21 | #include <linux/host1x.h> |
| 22 | #include <linux/kref.h> |
| 23 | #include <linux/module.h> |
| 24 | #include <linux/scatterlist.h> |
| 25 | #include <linux/slab.h> |
| 26 | #include <linux/vmalloc.h> |
| 27 | #include <trace/events/host1x.h> |
| 28 | |
| 29 | #include "channel.h" |
| 30 | #include "dev.h" |
| 31 | #include "job.h" |
| 32 | #include "syncpt.h" |
| 33 | |
| 34 | #define HOST1X_WAIT_SYNCPT_OFFSET 0x8 |
| 35 | |
| 36 | struct host1x_job *host1x_job_alloc(struct host1x_channel *ch, |
| 37 | u32 num_cmdbufs, u32 num_relocs, |
| 38 | u32 num_waitchks) |
| 39 | { |
| 40 | struct host1x_job *job = NULL; |
| 41 | unsigned int num_unpins = num_cmdbufs + num_relocs; |
| 42 | u64 total; |
| 43 | void *mem; |
| 44 | |
| 45 | /* Check that we're not going to overflow */ |
| 46 | total = sizeof(struct host1x_job) + |
| 47 | (u64)num_relocs * sizeof(struct host1x_reloc) + |
| 48 | (u64)num_unpins * sizeof(struct host1x_job_unpin_data) + |
| 49 | (u64)num_waitchks * sizeof(struct host1x_waitchk) + |
| 50 | (u64)num_cmdbufs * sizeof(struct host1x_job_gather) + |
| 51 | (u64)num_unpins * sizeof(dma_addr_t) + |
| 52 | (u64)num_unpins * sizeof(u32 *); |
| 53 | if (total > ULONG_MAX) |
| 54 | return NULL; |
| 55 | |
| 56 | mem = job = kzalloc(total, GFP_KERNEL); |
| 57 | if (!job) |
| 58 | return NULL; |
| 59 | |
| 60 | kref_init(&job->ref); |
| 61 | job->channel = ch; |
| 62 | |
| 63 | /* Redistribute memory to the structs */ |
| 64 | mem += sizeof(struct host1x_job); |
| 65 | job->relocarray = num_relocs ? mem : NULL; |
| 66 | mem += num_relocs * sizeof(struct host1x_reloc); |
| 67 | job->unpins = num_unpins ? mem : NULL; |
| 68 | mem += num_unpins * sizeof(struct host1x_job_unpin_data); |
| 69 | job->waitchk = num_waitchks ? mem : NULL; |
| 70 | mem += num_waitchks * sizeof(struct host1x_waitchk); |
| 71 | job->gathers = num_cmdbufs ? mem : NULL; |
| 72 | mem += num_cmdbufs * sizeof(struct host1x_job_gather); |
| 73 | job->addr_phys = num_unpins ? mem : NULL; |
| 74 | |
| 75 | job->reloc_addr_phys = job->addr_phys; |
| 76 | job->gather_addr_phys = &job->addr_phys[num_relocs]; |
| 77 | |
| 78 | return job; |
| 79 | } |
| 80 | EXPORT_SYMBOL(host1x_job_alloc); |
| 81 | |
| 82 | struct host1x_job *host1x_job_get(struct host1x_job *job) |
| 83 | { |
| 84 | kref_get(&job->ref); |
| 85 | return job; |
| 86 | } |
| 87 | EXPORT_SYMBOL(host1x_job_get); |
| 88 | |
| 89 | static void job_free(struct kref *ref) |
| 90 | { |
| 91 | struct host1x_job *job = container_of(ref, struct host1x_job, ref); |
| 92 | |
| 93 | kfree(job); |
| 94 | } |
| 95 | |
| 96 | void host1x_job_put(struct host1x_job *job) |
| 97 | { |
| 98 | kref_put(&job->ref, job_free); |
| 99 | } |
| 100 | EXPORT_SYMBOL(host1x_job_put); |
| 101 | |
| 102 | void host1x_job_add_gather(struct host1x_job *job, struct host1x_bo *bo, |
| 103 | u32 words, u32 offset) |
| 104 | { |
| 105 | struct host1x_job_gather *cur_gather = &job->gathers[job->num_gathers]; |
| 106 | |
| 107 | cur_gather->words = words; |
| 108 | cur_gather->bo = bo; |
| 109 | cur_gather->offset = offset; |
| 110 | job->num_gathers++; |
| 111 | } |
| 112 | EXPORT_SYMBOL(host1x_job_add_gather); |
| 113 | |
| 114 | /* |
| 115 | * NULL an already satisfied WAIT_SYNCPT host method, by patching its |
| 116 | * args in the command stream. The method data is changed to reference |
| 117 | * a reserved (never given out or incr) HOST1X_SYNCPT_RESERVED syncpt |
| 118 | * with a matching threshold value of 0, so is guaranteed to be popped |
| 119 | * by the host HW. |
| 120 | */ |
| 121 | static void host1x_syncpt_patch_offset(struct host1x_syncpt *sp, |
| 122 | struct host1x_bo *h, u32 offset) |
| 123 | { |
| 124 | void *patch_addr = NULL; |
| 125 | |
| 126 | /* patch the wait */ |
| 127 | patch_addr = host1x_bo_kmap(h, offset >> PAGE_SHIFT); |
| 128 | if (patch_addr) { |
| 129 | host1x_syncpt_patch_wait(sp, |
| 130 | patch_addr + (offset & ~PAGE_MASK)); |
| 131 | host1x_bo_kunmap(h, offset >> PAGE_SHIFT, patch_addr); |
| 132 | } else |
| 133 | pr_err("Could not map cmdbuf for wait check\n"); |
| 134 | } |
| 135 | |
| 136 | /* |
| 137 | * Check driver supplied waitchk structs for syncpt thresholds |
| 138 | * that have already been satisfied and NULL the comparison (to |
| 139 | * avoid a wrap condition in the HW). |
| 140 | */ |
| 141 | static int do_waitchks(struct host1x_job *job, struct host1x *host, |
| 142 | struct host1x_job_gather *g) |
| 143 | { |
| 144 | struct host1x_bo *patch = g->bo; |
| 145 | int i; |
| 146 | |
| 147 | /* compare syncpt vs wait threshold */ |
| 148 | for (i = 0; i < job->num_waitchk; i++) { |
| 149 | struct host1x_waitchk *wait = &job->waitchk[i]; |
| 150 | struct host1x_syncpt *sp = |
| 151 | host1x_syncpt_get(host, wait->syncpt_id); |
| 152 | |
| 153 | /* validate syncpt id */ |
| 154 | if (wait->syncpt_id > host1x_syncpt_nb_pts(host)) |
| 155 | continue; |
| 156 | |
| 157 | /* skip all other gathers */ |
| 158 | if (patch != wait->bo) |
| 159 | continue; |
| 160 | |
| 161 | trace_host1x_syncpt_wait_check(wait->bo, wait->offset, |
| 162 | wait->syncpt_id, wait->thresh, |
| 163 | host1x_syncpt_read_min(sp)); |
| 164 | |
| 165 | if (host1x_syncpt_is_expired(sp, wait->thresh)) { |
| 166 | dev_dbg(host->dev, |
| 167 | "drop WAIT id %u (%s) thresh 0x%x, min 0x%x\n", |
| 168 | wait->syncpt_id, sp->name, wait->thresh, |
| 169 | host1x_syncpt_read_min(sp)); |
| 170 | |
| 171 | host1x_syncpt_patch_offset(sp, patch, |
| 172 | g->offset + wait->offset); |
| 173 | } |
| 174 | |
| 175 | wait->bo = NULL; |
| 176 | } |
| 177 | |
| 178 | return 0; |
| 179 | } |
| 180 | |
| 181 | static unsigned int pin_job(struct host1x *host, struct host1x_job *job) |
| 182 | { |
| 183 | unsigned int i; |
| 184 | int err; |
| 185 | |
| 186 | job->num_unpins = 0; |
| 187 | |
| 188 | for (i = 0; i < job->num_relocs; i++) { |
| 189 | struct host1x_reloc *reloc = &job->relocarray[i]; |
| 190 | struct sg_table *sgt; |
| 191 | dma_addr_t phys_addr; |
| 192 | |
| 193 | reloc->target.bo = host1x_bo_get(reloc->target.bo); |
| 194 | if (!reloc->target.bo) { |
| 195 | err = -EINVAL; |
| 196 | goto unpin; |
| 197 | } |
| 198 | |
| 199 | phys_addr = host1x_bo_pin(reloc->target.bo, &sgt); |
| 200 | |
| 201 | job->addr_phys[job->num_unpins] = phys_addr; |
| 202 | job->unpins[job->num_unpins].bo = reloc->target.bo; |
| 203 | job->unpins[job->num_unpins].sgt = sgt; |
| 204 | job->num_unpins++; |
| 205 | } |
| 206 | |
| 207 | for (i = 0; i < job->num_gathers; i++) { |
| 208 | struct host1x_job_gather *g = &job->gathers[i]; |
| 209 | size_t gather_size = 0; |
| 210 | struct scatterlist *sg; |
| 211 | struct sg_table *sgt; |
| 212 | dma_addr_t phys_addr; |
| 213 | unsigned long shift; |
| 214 | struct iova *alloc; |
| 215 | unsigned int j; |
| 216 | |
| 217 | g->bo = host1x_bo_get(g->bo); |
| 218 | if (!g->bo) { |
| 219 | err = -EINVAL; |
| 220 | goto unpin; |
| 221 | } |
| 222 | |
| 223 | phys_addr = host1x_bo_pin(g->bo, &sgt); |
| 224 | |
| 225 | if (!IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && host->domain) { |
| 226 | for_each_sg(sgt->sgl, sg, sgt->nents, j) |
| 227 | gather_size += sg->length; |
| 228 | gather_size = iova_align(&host->iova, gather_size); |
| 229 | |
| 230 | shift = iova_shift(&host->iova); |
| 231 | alloc = alloc_iova(&host->iova, gather_size >> shift, |
| 232 | host->iova_end >> shift, true); |
| 233 | if (!alloc) { |
| 234 | err = -ENOMEM; |
| 235 | goto unpin; |
| 236 | } |
| 237 | |
| 238 | err = iommu_map_sg(host->domain, |
| 239 | iova_dma_addr(&host->iova, alloc), |
| 240 | sgt->sgl, sgt->nents, IOMMU_READ); |
| 241 | if (err == 0) { |
| 242 | __free_iova(&host->iova, alloc); |
| 243 | err = -EINVAL; |
| 244 | goto unpin; |
| 245 | } |
| 246 | |
| 247 | job->addr_phys[job->num_unpins] = |
| 248 | iova_dma_addr(&host->iova, alloc); |
| 249 | job->unpins[job->num_unpins].size = gather_size; |
| 250 | } else { |
| 251 | job->addr_phys[job->num_unpins] = phys_addr; |
| 252 | } |
| 253 | |
| 254 | job->gather_addr_phys[i] = job->addr_phys[job->num_unpins]; |
| 255 | |
| 256 | job->unpins[job->num_unpins].bo = g->bo; |
| 257 | job->unpins[job->num_unpins].sgt = sgt; |
| 258 | job->num_unpins++; |
| 259 | } |
| 260 | |
| 261 | return 0; |
| 262 | |
| 263 | unpin: |
| 264 | host1x_job_unpin(job); |
| 265 | return err; |
| 266 | } |
| 267 | |
| 268 | static int do_relocs(struct host1x_job *job, struct host1x_job_gather *g) |
| 269 | { |
| 270 | int i = 0; |
| 271 | u32 last_page = ~0; |
| 272 | void *cmdbuf_page_addr = NULL; |
| 273 | struct host1x_bo *cmdbuf = g->bo; |
| 274 | |
| 275 | /* pin & patch the relocs for one gather */ |
| 276 | for (i = 0; i < job->num_relocs; i++) { |
| 277 | struct host1x_reloc *reloc = &job->relocarray[i]; |
| 278 | u32 reloc_addr = (job->reloc_addr_phys[i] + |
| 279 | reloc->target.offset) >> reloc->shift; |
| 280 | u32 *target; |
| 281 | |
| 282 | /* skip all other gathers */ |
| 283 | if (cmdbuf != reloc->cmdbuf.bo) |
| 284 | continue; |
| 285 | |
| 286 | if (IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL)) { |
| 287 | target = (u32 *)job->gather_copy_mapped + |
| 288 | reloc->cmdbuf.offset / sizeof(u32) + |
| 289 | g->offset / sizeof(u32); |
| 290 | goto patch_reloc; |
| 291 | } |
| 292 | |
| 293 | if (last_page != reloc->cmdbuf.offset >> PAGE_SHIFT) { |
| 294 | if (cmdbuf_page_addr) |
| 295 | host1x_bo_kunmap(cmdbuf, last_page, |
| 296 | cmdbuf_page_addr); |
| 297 | |
| 298 | cmdbuf_page_addr = host1x_bo_kmap(cmdbuf, |
| 299 | reloc->cmdbuf.offset >> PAGE_SHIFT); |
| 300 | last_page = reloc->cmdbuf.offset >> PAGE_SHIFT; |
| 301 | |
| 302 | if (unlikely(!cmdbuf_page_addr)) { |
| 303 | pr_err("Could not map cmdbuf for relocation\n"); |
| 304 | return -ENOMEM; |
| 305 | } |
| 306 | } |
| 307 | |
| 308 | target = cmdbuf_page_addr + (reloc->cmdbuf.offset & ~PAGE_MASK); |
| 309 | patch_reloc: |
| 310 | *target = reloc_addr; |
| 311 | } |
| 312 | |
| 313 | if (cmdbuf_page_addr) |
| 314 | host1x_bo_kunmap(cmdbuf, last_page, cmdbuf_page_addr); |
| 315 | |
| 316 | return 0; |
| 317 | } |
| 318 | |
| 319 | static bool check_reloc(struct host1x_reloc *reloc, struct host1x_bo *cmdbuf, |
| 320 | unsigned int offset) |
| 321 | { |
| 322 | offset *= sizeof(u32); |
| 323 | |
| 324 | if (reloc->cmdbuf.bo != cmdbuf || reloc->cmdbuf.offset != offset) |
| 325 | return false; |
| 326 | |
| 327 | /* relocation shift value validation isn't implemented yet */ |
| 328 | if (reloc->shift) |
| 329 | return false; |
| 330 | |
| 331 | return true; |
| 332 | } |
| 333 | |
| 334 | static bool check_wait(struct host1x_waitchk *wait, struct host1x_bo *cmdbuf, |
| 335 | unsigned int offset) |
| 336 | { |
| 337 | offset *= sizeof(u32); |
| 338 | |
| 339 | if (wait->bo != cmdbuf || wait->offset != offset) |
| 340 | return false; |
| 341 | |
| 342 | return true; |
| 343 | } |
| 344 | |
| 345 | struct host1x_firewall { |
| 346 | struct host1x_job *job; |
| 347 | struct device *dev; |
| 348 | |
| 349 | unsigned int num_relocs; |
| 350 | struct host1x_reloc *reloc; |
| 351 | |
| 352 | unsigned int num_waitchks; |
| 353 | struct host1x_waitchk *waitchk; |
| 354 | |
| 355 | struct host1x_bo *cmdbuf; |
| 356 | unsigned int offset; |
| 357 | |
| 358 | u32 words; |
| 359 | u32 class; |
| 360 | u32 reg; |
| 361 | u32 mask; |
| 362 | u32 count; |
| 363 | }; |
| 364 | |
| 365 | static int check_register(struct host1x_firewall *fw, unsigned long offset) |
| 366 | { |
| 367 | if (!fw->job->is_addr_reg) |
| 368 | return 0; |
| 369 | |
| 370 | if (fw->job->is_addr_reg(fw->dev, fw->class, offset)) { |
| 371 | if (!fw->num_relocs) |
| 372 | return -EINVAL; |
| 373 | |
| 374 | if (!check_reloc(fw->reloc, fw->cmdbuf, fw->offset)) |
| 375 | return -EINVAL; |
| 376 | |
| 377 | fw->num_relocs--; |
| 378 | fw->reloc++; |
| 379 | } |
| 380 | |
| 381 | if (offset == HOST1X_WAIT_SYNCPT_OFFSET) { |
| 382 | if (fw->class != HOST1X_CLASS_HOST1X) |
| 383 | return -EINVAL; |
| 384 | |
| 385 | if (!fw->num_waitchks) |
| 386 | return -EINVAL; |
| 387 | |
| 388 | if (!check_wait(fw->waitchk, fw->cmdbuf, fw->offset)) |
| 389 | return -EINVAL; |
| 390 | |
| 391 | fw->num_waitchks--; |
| 392 | fw->waitchk++; |
| 393 | } |
| 394 | |
| 395 | return 0; |
| 396 | } |
| 397 | |
| 398 | static int check_class(struct host1x_firewall *fw, u32 class) |
| 399 | { |
| 400 | if (!fw->job->is_valid_class) { |
| 401 | if (fw->class != class) |
| 402 | return -EINVAL; |
| 403 | } else { |
| 404 | if (!fw->job->is_valid_class(fw->class)) |
| 405 | return -EINVAL; |
| 406 | } |
| 407 | |
| 408 | return 0; |
| 409 | } |
| 410 | |
| 411 | static int check_mask(struct host1x_firewall *fw) |
| 412 | { |
| 413 | u32 mask = fw->mask; |
| 414 | u32 reg = fw->reg; |
| 415 | int ret; |
| 416 | |
| 417 | while (mask) { |
| 418 | if (fw->words == 0) |
| 419 | return -EINVAL; |
| 420 | |
| 421 | if (mask & 1) { |
| 422 | ret = check_register(fw, reg); |
| 423 | if (ret < 0) |
| 424 | return ret; |
| 425 | |
| 426 | fw->words--; |
| 427 | fw->offset++; |
| 428 | } |
| 429 | mask >>= 1; |
| 430 | reg++; |
| 431 | } |
| 432 | |
| 433 | return 0; |
| 434 | } |
| 435 | |
| 436 | static int check_incr(struct host1x_firewall *fw) |
| 437 | { |
| 438 | u32 count = fw->count; |
| 439 | u32 reg = fw->reg; |
| 440 | int ret; |
| 441 | |
| 442 | while (count) { |
| 443 | if (fw->words == 0) |
| 444 | return -EINVAL; |
| 445 | |
| 446 | ret = check_register(fw, reg); |
| 447 | if (ret < 0) |
| 448 | return ret; |
| 449 | |
| 450 | reg++; |
| 451 | fw->words--; |
| 452 | fw->offset++; |
| 453 | count--; |
| 454 | } |
| 455 | |
| 456 | return 0; |
| 457 | } |
| 458 | |
| 459 | static int check_nonincr(struct host1x_firewall *fw) |
| 460 | { |
| 461 | u32 count = fw->count; |
| 462 | int ret; |
| 463 | |
| 464 | while (count) { |
| 465 | if (fw->words == 0) |
| 466 | return -EINVAL; |
| 467 | |
| 468 | ret = check_register(fw, fw->reg); |
| 469 | if (ret < 0) |
| 470 | return ret; |
| 471 | |
| 472 | fw->words--; |
| 473 | fw->offset++; |
| 474 | count--; |
| 475 | } |
| 476 | |
| 477 | return 0; |
| 478 | } |
| 479 | |
| 480 | static int validate(struct host1x_firewall *fw, struct host1x_job_gather *g) |
| 481 | { |
| 482 | u32 *cmdbuf_base = (u32 *)fw->job->gather_copy_mapped + |
| 483 | (g->offset / sizeof(u32)); |
| 484 | u32 job_class = fw->class; |
| 485 | int err = 0; |
| 486 | |
| 487 | fw->words = g->words; |
| 488 | fw->cmdbuf = g->bo; |
| 489 | fw->offset = 0; |
| 490 | |
| 491 | while (fw->words && !err) { |
| 492 | u32 word = cmdbuf_base[fw->offset]; |
| 493 | u32 opcode = (word & 0xf0000000) >> 28; |
| 494 | |
| 495 | fw->mask = 0; |
| 496 | fw->reg = 0; |
| 497 | fw->count = 0; |
| 498 | fw->words--; |
| 499 | fw->offset++; |
| 500 | |
| 501 | switch (opcode) { |
| 502 | case 0: |
| 503 | fw->class = word >> 6 & 0x3ff; |
| 504 | fw->mask = word & 0x3f; |
| 505 | fw->reg = word >> 16 & 0xfff; |
| 506 | err = check_class(fw, job_class); |
| 507 | if (!err) |
| 508 | err = check_mask(fw); |
| 509 | if (err) |
| 510 | goto out; |
| 511 | break; |
| 512 | case 1: |
| 513 | fw->reg = word >> 16 & 0xfff; |
| 514 | fw->count = word & 0xffff; |
| 515 | err = check_incr(fw); |
| 516 | if (err) |
| 517 | goto out; |
| 518 | break; |
| 519 | |
| 520 | case 2: |
| 521 | fw->reg = word >> 16 & 0xfff; |
| 522 | fw->count = word & 0xffff; |
| 523 | err = check_nonincr(fw); |
| 524 | if (err) |
| 525 | goto out; |
| 526 | break; |
| 527 | |
| 528 | case 3: |
| 529 | fw->mask = word & 0xffff; |
| 530 | fw->reg = word >> 16 & 0xfff; |
| 531 | err = check_mask(fw); |
| 532 | if (err) |
| 533 | goto out; |
| 534 | break; |
| 535 | case 4: |
| 536 | case 14: |
| 537 | break; |
| 538 | default: |
| 539 | err = -EINVAL; |
| 540 | break; |
| 541 | } |
| 542 | } |
| 543 | |
| 544 | out: |
| 545 | return err; |
| 546 | } |
| 547 | |
| 548 | static inline int copy_gathers(struct device *host, struct host1x_job *job, |
| 549 | struct device *dev) |
| 550 | { |
| 551 | struct host1x_firewall fw; |
| 552 | size_t size = 0; |
| 553 | size_t offset = 0; |
| 554 | int i; |
| 555 | |
| 556 | fw.job = job; |
| 557 | fw.dev = dev; |
| 558 | fw.reloc = job->relocarray; |
| 559 | fw.num_relocs = job->num_relocs; |
| 560 | fw.waitchk = job->waitchk; |
| 561 | fw.num_waitchks = job->num_waitchk; |
| 562 | fw.class = job->class; |
| 563 | |
| 564 | for (i = 0; i < job->num_gathers; i++) { |
| 565 | struct host1x_job_gather *g = &job->gathers[i]; |
| 566 | |
| 567 | size += g->words * sizeof(u32); |
| 568 | } |
| 569 | |
| 570 | /* |
| 571 | * Try a non-blocking allocation from a higher priority pools first, |
| 572 | * as awaiting for the allocation here is a major performance hit. |
| 573 | */ |
| 574 | job->gather_copy_mapped = dma_alloc_wc(host, size, &job->gather_copy, |
| 575 | GFP_NOWAIT); |
| 576 | |
| 577 | /* the higher priority allocation failed, try the generic-blocking */ |
| 578 | if (!job->gather_copy_mapped) |
| 579 | job->gather_copy_mapped = dma_alloc_wc(host, size, |
| 580 | &job->gather_copy, |
| 581 | GFP_KERNEL); |
| 582 | if (!job->gather_copy_mapped) |
| 583 | return -ENOMEM; |
| 584 | |
| 585 | job->gather_copy_size = size; |
| 586 | |
| 587 | for (i = 0; i < job->num_gathers; i++) { |
| 588 | struct host1x_job_gather *g = &job->gathers[i]; |
| 589 | void *gather; |
| 590 | |
| 591 | /* Copy the gather */ |
| 592 | gather = host1x_bo_mmap(g->bo); |
| 593 | memcpy(job->gather_copy_mapped + offset, gather + g->offset, |
| 594 | g->words * sizeof(u32)); |
| 595 | host1x_bo_munmap(g->bo, gather); |
| 596 | |
| 597 | /* Store the location in the buffer */ |
| 598 | g->base = job->gather_copy; |
| 599 | g->offset = offset; |
| 600 | |
| 601 | /* Validate the job */ |
| 602 | if (validate(&fw, g)) |
| 603 | return -EINVAL; |
| 604 | |
| 605 | offset += g->words * sizeof(u32); |
| 606 | } |
| 607 | |
| 608 | /* No relocs and waitchks should remain at this point */ |
| 609 | if (fw.num_relocs || fw.num_waitchks) |
| 610 | return -EINVAL; |
| 611 | |
| 612 | return 0; |
| 613 | } |
| 614 | |
| 615 | int host1x_job_pin(struct host1x_job *job, struct device *dev) |
| 616 | { |
| 617 | int err; |
| 618 | unsigned int i, j; |
| 619 | struct host1x *host = dev_get_drvdata(dev->parent); |
| 620 | DECLARE_BITMAP(waitchk_mask, host1x_syncpt_nb_pts(host)); |
| 621 | |
| 622 | bitmap_zero(waitchk_mask, host1x_syncpt_nb_pts(host)); |
| 623 | for (i = 0; i < job->num_waitchk; i++) { |
| 624 | u32 syncpt_id = job->waitchk[i].syncpt_id; |
| 625 | |
| 626 | if (syncpt_id < host1x_syncpt_nb_pts(host)) |
| 627 | set_bit(syncpt_id, waitchk_mask); |
| 628 | } |
| 629 | |
| 630 | /* get current syncpt values for waitchk */ |
| 631 | for_each_set_bit(i, waitchk_mask, host1x_syncpt_nb_pts(host)) |
| 632 | host1x_syncpt_load(host->syncpt + i); |
| 633 | |
| 634 | /* pin memory */ |
| 635 | err = pin_job(host, job); |
| 636 | if (err) |
| 637 | goto out; |
| 638 | |
| 639 | if (IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL)) { |
| 640 | err = copy_gathers(host->dev, job, dev); |
| 641 | if (err) |
| 642 | goto out; |
| 643 | } |
| 644 | |
| 645 | /* patch gathers */ |
| 646 | for (i = 0; i < job->num_gathers; i++) { |
| 647 | struct host1x_job_gather *g = &job->gathers[i]; |
| 648 | |
| 649 | /* process each gather mem only once */ |
| 650 | if (g->handled) |
| 651 | continue; |
| 652 | |
| 653 | /* copy_gathers() sets gathers base if firewall is enabled */ |
| 654 | if (!IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL)) |
| 655 | g->base = job->gather_addr_phys[i]; |
| 656 | |
| 657 | for (j = i + 1; j < job->num_gathers; j++) { |
| 658 | if (job->gathers[j].bo == g->bo) { |
| 659 | job->gathers[j].handled = true; |
| 660 | job->gathers[j].base = g->base; |
| 661 | } |
| 662 | } |
| 663 | |
| 664 | err = do_relocs(job, g); |
| 665 | if (err) |
| 666 | break; |
| 667 | |
| 668 | err = do_waitchks(job, host, g); |
| 669 | if (err) |
| 670 | break; |
| 671 | } |
| 672 | |
| 673 | out: |
| 674 | if (err) |
| 675 | host1x_job_unpin(job); |
| 676 | wmb(); |
| 677 | |
| 678 | return err; |
| 679 | } |
| 680 | EXPORT_SYMBOL(host1x_job_pin); |
| 681 | |
| 682 | void host1x_job_unpin(struct host1x_job *job) |
| 683 | { |
| 684 | struct host1x *host = dev_get_drvdata(job->channel->dev->parent); |
| 685 | unsigned int i; |
| 686 | |
| 687 | for (i = 0; i < job->num_unpins; i++) { |
| 688 | struct host1x_job_unpin_data *unpin = &job->unpins[i]; |
| 689 | |
| 690 | if (!IS_ENABLED(CONFIG_TEGRA_HOST1X_FIREWALL) && |
| 691 | unpin->size && host->domain) { |
| 692 | iommu_unmap(host->domain, job->addr_phys[i], |
| 693 | unpin->size); |
| 694 | free_iova(&host->iova, |
| 695 | iova_pfn(&host->iova, job->addr_phys[i])); |
| 696 | } |
| 697 | |
| 698 | host1x_bo_unpin(unpin->bo, unpin->sgt); |
| 699 | host1x_bo_put(unpin->bo); |
| 700 | } |
| 701 | |
| 702 | job->num_unpins = 0; |
| 703 | |
| 704 | if (job->gather_copy_size) |
| 705 | dma_free_wc(host->dev, job->gather_copy_size, |
| 706 | job->gather_copy_mapped, job->gather_copy); |
| 707 | } |
| 708 | EXPORT_SYMBOL(host1x_job_unpin); |
| 709 | |
| 710 | /* |
| 711 | * Debug routine used to dump job entries |
| 712 | */ |
| 713 | void host1x_job_dump(struct device *dev, struct host1x_job *job) |
| 714 | { |
| 715 | dev_dbg(dev, " SYNCPT_ID %d\n", job->syncpt_id); |
| 716 | dev_dbg(dev, " SYNCPT_VAL %d\n", job->syncpt_end); |
| 717 | dev_dbg(dev, " FIRST_GET 0x%x\n", job->first_get); |
| 718 | dev_dbg(dev, " TIMEOUT %d\n", job->timeout); |
| 719 | dev_dbg(dev, " NUM_SLOTS %d\n", job->num_slots); |
| 720 | dev_dbg(dev, " NUM_HANDLES %d\n", job->num_unpins); |
| 721 | } |