ctxgm200.c 4.2 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136
  1. /*
  2. * Copyright 2015 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 "ctxgf100.h"
  25. /*******************************************************************************
  26. * PGRAPH context implementation
  27. ******************************************************************************/
  28. void
  29. gm200_grctx_generate_tpcid(struct gf100_gr *gr)
  30. {
  31. struct nvkm_device *device = gr->base.engine.subdev.device;
  32. int gpc, tpc, id;
  33. for (tpc = 0, id = 0; tpc < TPC_MAX_PER_GPC; tpc++) {
  34. for (gpc = 0; gpc < gr->gpc_nr; gpc++) {
  35. if (tpc < gr->tpc_nr[gpc]) {
  36. nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x698), id);
  37. nvkm_wr32(device, GPC_UNIT(gpc, 0x0c10 + tpc * 4), id);
  38. nvkm_wr32(device, TPC_UNIT(gpc, tpc, 0x088), id);
  39. id++;
  40. }
  41. }
  42. }
  43. }
  44. void
  45. gm200_grctx_generate_405b60(struct gf100_gr *gr)
  46. {
  47. struct nvkm_device *device = gr->base.engine.subdev.device;
  48. const u32 dist_nr = DIV_ROUND_UP(gr->tpc_total, 4);
  49. u32 dist[TPC_MAX / 4] = {};
  50. u32 gpcs[GPC_MAX] = {};
  51. u8 tpcnr[GPC_MAX];
  52. int tpc, gpc, i;
  53. memcpy(tpcnr, gr->tpc_nr, sizeof(gr->tpc_nr));
  54. /* won't result in the same distribution as the binary driver where
  55. * some of the gpcs have more tpcs than others, but this shall do
  56. * for the moment. the code for earlier gpus has this issue too.
  57. */
  58. for (gpc = -1, i = 0; i < gr->tpc_total; i++) {
  59. do {
  60. gpc = (gpc + 1) % gr->gpc_nr;
  61. } while(!tpcnr[gpc]);
  62. tpc = gr->tpc_nr[gpc] - tpcnr[gpc]--;
  63. dist[i / 4] |= ((gpc << 4) | tpc) << ((i % 4) * 8);
  64. gpcs[gpc] |= i << (tpc * 8);
  65. }
  66. for (i = 0; i < dist_nr; i++)
  67. nvkm_wr32(device, 0x405b60 + (i * 4), dist[i]);
  68. for (i = 0; i < gr->gpc_nr; i++)
  69. nvkm_wr32(device, 0x405ba0 + (i * 4), gpcs[i]);
  70. }
  71. static void
  72. gm200_grctx_generate_main(struct gf100_gr *gr, struct gf100_grctx *info)
  73. {
  74. struct nvkm_device *device = gr->base.engine.subdev.device;
  75. const struct gf100_grctx_func *grctx = gr->func->grctx;
  76. u32 idle_timeout, tmp;
  77. int i;
  78. gf100_gr_mmio(gr, gr->fuc_sw_ctx);
  79. idle_timeout = nvkm_mask(device, 0x404154, 0xffffffff, 0x00000000);
  80. grctx->bundle(info);
  81. grctx->pagepool(info);
  82. grctx->attrib(info);
  83. grctx->unkn(gr);
  84. gm200_grctx_generate_tpcid(gr);
  85. gf100_grctx_generate_r406028(gr);
  86. gk104_grctx_generate_r418bb8(gr);
  87. for (i = 0; i < 8; i++)
  88. nvkm_wr32(device, 0x4064d0 + (i * 0x04), 0x00000000);
  89. nvkm_wr32(device, 0x406500, 0x00000000);
  90. nvkm_wr32(device, 0x405b00, (gr->tpc_total << 8) | gr->gpc_nr);
  91. for (tmp = 0, i = 0; i < gr->gpc_nr; i++)
  92. tmp |= ((1 << gr->tpc_nr[i]) - 1) << (i * 4);
  93. nvkm_wr32(device, 0x4041c4, tmp);
  94. gm200_grctx_generate_405b60(gr);
  95. gf100_gr_icmd(gr, gr->fuc_bundle);
  96. nvkm_wr32(device, 0x404154, idle_timeout);
  97. gf100_gr_mthd(gr, gr->fuc_method);
  98. nvkm_mask(device, 0x418e94, 0xffffffff, 0xc4230000);
  99. nvkm_mask(device, 0x418e4c, 0xffffffff, 0x70000000);
  100. }
  101. const struct gf100_grctx_func
  102. gm200_grctx = {
  103. .main = gm200_grctx_generate_main,
  104. .unkn = gk104_grctx_generate_unkn,
  105. .bundle = gm107_grctx_generate_bundle,
  106. .bundle_size = 0x3000,
  107. .bundle_min_gpm_fifo_depth = 0x180,
  108. .bundle_token_limit = 0x780,
  109. .pagepool = gm107_grctx_generate_pagepool,
  110. .pagepool_size = 0x20000,
  111. .attrib = gm107_grctx_generate_attrib,
  112. .attrib_nr_max = 0x600,
  113. .attrib_nr = 0x400,
  114. .alpha_nr_max = 0x1800,
  115. .alpha_nr = 0x1000,
  116. };