diff options
Diffstat (limited to 'drivers/gpu/drm/i915')
34 files changed, 982 insertions, 508 deletions
diff --git a/drivers/gpu/drm/i915/gvt/aperture_gm.c b/drivers/gpu/drm/i915/gvt/aperture_gm.c index fe754022e356..359d37d5c958 100644 --- a/drivers/gpu/drm/i915/gvt/aperture_gm.c +++ b/drivers/gpu/drm/i915/gvt/aperture_gm.c @@ -61,10 +61,12 @@ static int alloc_gm(struct intel_vgpu *vgpu, bool high_gm) } mutex_lock(&dev_priv->drm.struct_mutex); + mmio_hw_access_pre(dev_priv); ret = i915_gem_gtt_insert(&dev_priv->ggtt.vm, node, size, I915_GTT_PAGE_SIZE, I915_COLOR_UNEVICTABLE, start, end, flags); + mmio_hw_access_post(dev_priv); mutex_unlock(&dev_priv->drm.struct_mutex); if (ret) gvt_err("fail to alloc %s gm space from host\n", diff --git a/drivers/gpu/drm/i915/gvt/fb_decoder.c b/drivers/gpu/drm/i915/gvt/fb_decoder.c index 481896fb712a..85e6736f0a32 100644 --- a/drivers/gpu/drm/i915/gvt/fb_decoder.c +++ b/drivers/gpu/drm/i915/gvt/fb_decoder.c @@ -235,7 +235,7 @@ int intel_vgpu_decode_primary_plane(struct intel_vgpu *vgpu, plane->bpp = skl_pixel_formats[fmt].bpp; plane->drm_format = skl_pixel_formats[fmt].drm_format; } else { - plane->tiled = !!(val & DISPPLANE_TILED); + plane->tiled = val & DISPPLANE_TILED; fmt = bdw_format_to_drm(val & DISPPLANE_PIXFORMAT_MASK); plane->bpp = bdw_pixel_formats[fmt].bpp; plane->drm_format = bdw_pixel_formats[fmt].drm_format; diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c index 2402395a068d..c7103dd2d8d5 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.c +++ b/drivers/gpu/drm/i915/gvt/gtt.c @@ -1905,7 +1905,6 @@ static struct intel_vgpu_mm *intel_vgpu_create_ggtt_mm(struct intel_vgpu *vgpu) vgpu_free_mm(mm); return ERR_PTR(-ENOMEM); } - mm->ggtt_mm.last_partial_off = -1UL; return mm; } @@ -1930,7 +1929,6 @@ void _intel_vgpu_mm_release(struct kref *mm_ref) invalidate_ppgtt_mm(mm); } else { vfree(mm->ggtt_mm.virtual_ggtt); - mm->ggtt_mm.last_partial_off = -1UL; } vgpu_free_mm(mm); @@ -2168,6 +2166,8 @@ static int emulate_ggtt_mmio_write(struct intel_vgpu *vgpu, unsigned int off, struct intel_gvt_gtt_entry e, m; dma_addr_t dma_addr; int ret; + struct intel_gvt_partial_pte *partial_pte, *pos, *n; + bool partial_update = false; if (bytes != 4 && bytes != 8) return -EINVAL; @@ -2178,68 +2178,57 @@ static int emulate_ggtt_mmio_write(struct intel_vgpu *vgpu, unsigned int off, if (!vgpu_gmadr_is_valid(vgpu, gma)) return 0; - ggtt_get_guest_entry(ggtt_mm, &e, g_gtt_index); - + e.type = GTT_TYPE_GGTT_PTE; memcpy((void *)&e.val64 + (off & (info->gtt_entry_size - 1)), p_data, bytes); /* If ggtt entry size is 8 bytes, and it's split into two 4 bytes - * write, we assume the two 4 bytes writes are consecutive. - * Otherwise, we abort and report error + * write, save the first 4 bytes in a list and update virtual + * PTE. Only update shadow PTE when the second 4 bytes comes. */ if (bytes < info->gtt_entry_size) { - if (ggtt_mm->ggtt_mm.last_partial_off == -1UL) { - /* the first partial part*/ - ggtt_mm->ggtt_mm.last_partial_off = off; - ggtt_mm->ggtt_mm.last_partial_data = e.val64; - return 0; - } else if ((g_gtt_index == - (ggtt_mm->ggtt_mm.last_partial_off >> - info->gtt_entry_size_shift)) && - (off != ggtt_mm->ggtt_mm.last_partial_off)) { - /* the second partial part */ - - int last_off = ggtt_mm->ggtt_mm.last_partial_off & - (info->gtt_entry_size - 1); - - memcpy((void *)&e.val64 + last_off, - (void *)&ggtt_mm->ggtt_mm.last_partial_data + - last_off, bytes); - - ggtt_mm->ggtt_mm.last_partial_off = -1UL; - } else { - int last_offset; - - gvt_vgpu_err("failed to populate guest ggtt entry: abnormal ggtt entry write sequence, last_partial_off=%lx, offset=%x, bytes=%d, ggtt entry size=%d\n", - ggtt_mm->ggtt_mm.last_partial_off, off, - bytes, info->gtt_entry_size); - - /* set host ggtt entry to scratch page and clear - * virtual ggtt entry as not present for last - * partially write offset - */ - last_offset = ggtt_mm->ggtt_mm.last_partial_off & - (~(info->gtt_entry_size - 1)); - - ggtt_get_host_entry(ggtt_mm, &m, last_offset); - ggtt_invalidate_pte(vgpu, &m); - ops->set_pfn(&m, gvt->gtt.scratch_mfn); - ops->clear_present(&m); - ggtt_set_host_entry(ggtt_mm, &m, last_offset); - ggtt_invalidate(gvt->dev_priv); - - ggtt_get_guest_entry(ggtt_mm, &e, last_offset); - ops->clear_present(&e); - ggtt_set_guest_entry(ggtt_mm, &e, last_offset); - - ggtt_mm->ggtt_mm.last_partial_off = off; - ggtt_mm->ggtt_mm.last_partial_data = e.val64; + bool found = false; + + list_for_each_entry_safe(pos, n, + &ggtt_mm->ggtt_mm.partial_pte_list, list) { + if (g_gtt_index == pos->offset >> + info->gtt_entry_size_shift) { + if (off != pos->offset) { + /* the second partial part*/ + int last_off = pos->offset & + (info->gtt_entry_size - 1); + + memcpy((void *)&e.val64 + last_off, + (void *)&pos->data + last_off, + bytes); + + list_del(&pos->list); + kfree(pos); + found = true; + break; + } + + /* update of the first partial part */ + pos->data = e.val64; + ggtt_set_guest_entry(ggtt_mm, &e, g_gtt_index); + return 0; + } + } - return 0; + if (!found) { + /* the first partial part */ + partial_pte = kzalloc(sizeof(*partial_pte), GFP_KERNEL); + if (!partial_pte) + return -ENOMEM; + partial_pte->offset = off; + partial_pte->data = e.val64; + list_add_tail(&partial_pte->list, + &ggtt_mm->ggtt_mm.partial_pte_list); + partial_update = true; } } - if (ops->test_present(&e)) { + if (!partial_update && (ops->test_present(&e))) { gfn = ops->get_pfn(&e); m = e; @@ -2263,16 +2252,18 @@ static int emulate_ggtt_mmio_write(struct intel_vgpu *vgpu, unsigned int off, } else ops->set_pfn(&m, dma_addr >> PAGE_SHIFT); } else { - ggtt_get_host_entry(ggtt_mm, &m, g_gtt_index); - ggtt_invalidate_pte(vgpu, &m); ops->set_pfn(&m, gvt->gtt.scratch_mfn); ops->clear_present(&m); } out: + ggtt_set_guest_entry(ggtt_mm, &e, g_gtt_index); + + ggtt_get_host_entry(ggtt_mm, &e, g_gtt_index); + ggtt_invalidate_pte(vgpu, &e); + ggtt_set_host_entry(ggtt_mm, &m, g_gtt_index); ggtt_invalidate(gvt->dev_priv); - ggtt_set_guest_entry(ggtt_mm, &e, g_gtt_index); return 0; } @@ -2430,6 +2421,8 @@ int intel_vgpu_init_gtt(struct intel_vgpu *vgpu) intel_vgpu_reset_ggtt(vgpu, false); + INIT_LIST_HEAD(>t->ggtt_mm->ggtt_mm.partial_pte_list); + return create_scratch_page_tree(vgpu); } @@ -2454,6 +2447,15 @@ static void intel_vgpu_destroy_all_ppgtt_mm(struct intel_vgpu *vgpu) static void intel_vgpu_destroy_ggtt_mm(struct intel_vgpu *vgpu) { + struct intel_gvt_partial_pte *pos, *next; + + list_for_each_entry_safe(pos, next, + &vgpu->gtt.ggtt_mm->ggtt_mm.partial_pte_list, + list) { + gvt_dbg_mm("partial PTE update on hold 0x%lx : 0x%llx\n", + pos->offset, pos->data); + kfree(pos); + } intel_vgpu_destroy_mm(vgpu->gtt.ggtt_mm); vgpu->gtt.ggtt_mm = NULL; } diff --git a/drivers/gpu/drm/i915/gvt/gtt.h b/drivers/gpu/drm/i915/gvt/gtt.h index 7a9b36176efb..d8cb04cc946d 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.h +++ b/drivers/gpu/drm/i915/gvt/gtt.h @@ -35,7 +35,6 @@ #define _GVT_GTT_H_ #define I915_GTT_PAGE_SHIFT 12 -#define I915_GTT_PAGE_MASK (~(I915_GTT_PAGE_SIZE - 1)) struct intel_vgpu_mm; @@ -133,6 +132,12 @@ enum intel_gvt_mm_type { #define GVT_RING_CTX_NR_PDPS GEN8_3LVL_PDPES +struct intel_gvt_partial_pte { + unsigned long offset; + u64 data; + struct list_head list; +}; + struct intel_vgpu_mm { enum intel_gvt_mm_type type; struct intel_vgpu *vgpu; @@ -157,8 +162,7 @@ struct intel_vgpu_mm { } ppgtt_mm; struct { void *virtual_ggtt; - unsigned long last_partial_off; - u64 last_partial_data; + struct list_head partial_pte_list; } ggtt_mm; }; }; diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c index 90f50f67909a..aa280bb07125 100644 --- a/drivers/gpu/drm/i915/gvt/handlers.c +++ b/drivers/gpu/drm/i915/gvt/handlers.c @@ -1609,7 +1609,7 @@ static int bxt_gt_disp_pwron_write(struct intel_vgpu *vgpu, return 0; } -static int bxt_edp_psr_imr_iir_write(struct intel_vgpu *vgpu, +static int edp_psr_imr_iir_write(struct intel_vgpu *vgpu, unsigned int offset, void *p_data, unsigned int bytes) { vgpu_vreg(vgpu, offset) = 0; @@ -2607,6 +2607,9 @@ static int init_generic_mmio_info(struct intel_gvt *gvt) MMIO_DFH(_MMIO(0x1a178), D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL); MMIO_DFH(_MMIO(0x1a17c), D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL); MMIO_DFH(_MMIO(0x2217c), D_BDW_PLUS, F_CMD_ACCESS, NULL, NULL); + + MMIO_DH(EDP_PSR_IMR, D_BDW_PLUS, NULL, edp_psr_imr_iir_write); + MMIO_DH(EDP_PSR_IIR, D_BDW_PLUS, NULL, edp_psr_imr_iir_write); return 0; } @@ -3205,9 +3208,6 @@ static int init_bxt_mmio_info(struct intel_gvt *gvt) MMIO_D(HSW_TVIDEO_DIP_GCP(TRANSCODER_B), D_BXT); MMIO_D(HSW_TVIDEO_DIP_GCP(TRANSCODER_C), D_BXT); - MMIO_DH(EDP_PSR_IMR, D_BXT, NULL, bxt_edp_psr_imr_iir_write); - MMIO_DH(EDP_PSR_IIR, D_BXT, NULL, bxt_edp_psr_imr_iir_write); - MMIO_D(RC6_CTX_BASE, D_BXT); MMIO_D(GEN8_PUSHBUS_CONTROL, D_BXT); diff --git a/drivers/gpu/drm/i915/gvt/mmio_context.c b/drivers/gpu/drm/i915/gvt/mmio_context.c index 10e63eea5492..d6e02c15ef97 100644 --- a/drivers/gpu/drm/i915/gvt/mmio_context.c +++ b/drivers/gpu/drm/i915/gvt/mmio_context.c @@ -131,7 +131,7 @@ static struct engine_mmio gen9_engine_mmio_list[] __cacheline_aligned = { {RCS, GAMT_CHKN_BIT_REG, 0x0, false}, /* 0x4ab8 */ {RCS, GEN9_GAMT_ECO_REG_RW_IA, 0x0, false}, /* 0x4ab0 */ - {RCS, GEN9_CSFE_CHICKEN1_RCS, 0x0, false}, /* 0x20d4 */ + {RCS, GEN9_CSFE_CHICKEN1_RCS, 0xffff, false}, /* 0x20d4 */ {RCS, GEN8_GARBCNTL, 0x0, false}, /* 0xb004 */ {RCS, GEN7_FF_THREAD_MODE, 0x0, false}, /* 0x20a0 */ @@ -158,6 +158,8 @@ static void load_render_mocs(struct drm_i915_private *dev_priv) int ring_id, i; for (ring_id = 0; ring_id < ARRAY_SIZE(regs); ring_id++) { + if (!HAS_ENGINE(dev_priv, ring_id)) + continue; offset.reg = regs[ring_id]; for (i = 0; i < GEN9_MOCS_SIZE; i++) { gen9_render_mocs.control_table[ring_id][i] = diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 44e2c0f5ec50..47062ee979cf 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -1175,8 +1175,6 @@ skl_dram_get_channels_info(struct drm_i915_private *dev_priv) return -EINVAL; } - dram_info->valid_dimm = true; - /* * If any of the channel is single rank channel, worst case output * will be same as if single rank memory, so consider single rank @@ -1193,8 +1191,7 @@ skl_dram_get_channels_info(struct drm_i915_private *dev_priv) return -EINVAL; } - if (ch0.is_16gb_dimm || ch1.is_16gb_dimm) - dram_info->is_16gb_dimm = true; + dram_info->is_16gb_dimm = ch0.is_16gb_dimm || ch1.is_16gb_dimm; dev_priv->dram_info.symmetric_memory = intel_is_dram_symmetric(val_ch0, val_ch1, @@ -1314,7 +1311,6 @@ bxt_get_dram_info(struct drm_i915_private *dev_priv) return -EINVAL; } - dram_info->valid_dimm = true; dram_info->valid = true; return 0; } @@ -1327,12 +1323,17 @@ intel_get_dram_info(struct drm_i915_private *dev_priv) int ret; dram_info->valid = false; - dram_info->valid_dimm = false; - dram_info->is_16gb_dimm = false; dram_info->rank = I915_DRAM_RANK_INVALID; dram_info->bandwidth_kbps = 0; dram_info->num_channels = 0; + /* + * Assume 16Gb DIMMs are present until proven otherwise. + * This is only used for the level 0 watermark latency + * w/a which does not apply to bxt/glk. + */ + dram_info->is_16gb_dimm = !IS_GEN9_LP(dev_priv); + if (INTEL_GEN(dev_priv) < 9 || IS_GEMINILAKE(dev_priv)) return; @@ -1443,6 +1444,7 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) intel_uncore_sanitize(dev_priv); + intel_gt_init_workarounds(dev_priv); i915_gem_load_init_fences(dev_priv); /* On the 945G/GM, the chipset reports the MSI capability on the diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 8624b4bdc242..872a2e159a5f 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -67,6 +67,7 @@ #include "intel_ringbuffer.h" #include "intel_uncore.h" #include "intel_wopcm.h" +#include "intel_workarounds.h" #include "intel_uc.h" #include "i915_gem.h" @@ -1805,6 +1806,7 @@ struct drm_i915_private { int dpio_phy_iosf_port[I915_NUM_PHYS_VLV]; struct i915_workarounds workarounds; + struct i915_wa_list gt_wa_list; struct i915_frontbuffer_tracking fb_tracking; @@ -1948,7 +1950,6 @@ struct drm_i915_private { struct dram_info { bool valid; - bool valid_dimm; bool is_16gb_dimm; u8 num_channels; enum dram_rank { @@ -2149,6 +2150,8 @@ struct drm_i915_private { struct delayed_work idle_work; ktime_t last_init_time; + + struct i915_vma *scratch; } gt; /* perform PHY state sanity checks? */ @@ -3871,4 +3874,9 @@ static inline int intel_hws_csb_write_index(struct drm_i915_private *i915) return I915_HWS_CSB_WRITE_INDEX; } +static inline u32 i915_scratch_offset(const struct drm_i915_private *i915) +{ + return i915_ggtt_offset(i915->gt.scratch); +} + #endif diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 0c8aa57ce83b..6ae9a6080cc8 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -5305,7 +5305,7 @@ int i915_gem_init_hw(struct drm_i915_private *dev_priv) } } - intel_gt_workarounds_apply(dev_priv); + intel_gt_apply_workarounds(dev_priv); i915_gem_init_swizzling(dev_priv); @@ -5500,6 +5500,44 @@ err_active: goto out_ctx; } +static int +i915_gem_init_scratch(struct drm_i915_private *i915, unsigned int size) +{ + struct drm_i915_gem_object *obj; + struct i915_vma *vma; + int ret; + + obj = i915_gem_object_create_stolen(i915, size); + if (!obj) + obj = i915_gem_object_create_internal(i915, size); + if (IS_ERR(obj)) { + DRM_ERROR("Failed to allocate scratch page\n"); + return PTR_ERR(obj); + } + + vma = i915_vma_instance(obj, &i915->ggtt.vm, NULL); + if (IS_ERR(vma)) { + ret = PTR_ERR(vma); + goto err_unref; + } + + ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH); + if (ret) + goto err_unref; + + i915->gt.scratch = vma; + return 0; + +err_unref: + i915_gem_object_put(obj); + return ret; +} + +static void i915_gem_fini_scratch(struct drm_i915_private *i915) +{ + i915_vma_unpin_and_release(&i915->gt.scratch, 0); +} + int i915_gem_init(struct drm_i915_private *dev_priv) { int ret; @@ -5546,12 +5584,19 @@ int i915_gem_init(struct drm_i915_private *dev_priv) goto err_unlock; } - ret = i915_gem_contexts_init(dev_priv); + ret = i915_gem_init_scratch(dev_priv, + IS_GEN2(dev_priv) ? SZ_256K : PAGE_SIZE); if (ret) { GEM_BUG_ON(ret == -EIO); goto err_ggtt; } + ret = i915_gem_contexts_init(dev_priv); + if (ret) { + GEM_BUG_ON(ret == -EIO); + goto err_scratch; + } + ret = intel_engines_init(dev_priv); if (ret) { GEM_BUG_ON(ret == -EIO); @@ -5624,6 +5669,8 @@ err_pm: err_context: if (ret != -EIO) i915_gem_contexts_fini(dev_priv); +err_scratch: + i915_gem_fini_scratch(dev_priv); err_ggtt: err_unlock: intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); @@ -5675,8 +5722,11 @@ void i915_gem_fini(struct drm_i915_private *dev_priv) intel_uc_fini(dev_priv); i915_gem_cleanup_engines(dev_priv); i915_gem_contexts_fini(dev_priv); + i915_gem_fini_scratch(dev_priv); mutex_unlock(&dev_priv->drm.struct_mutex); + intel_wa_list_free(&dev_priv->gt_wa_list); + intel_cleanup_gt_powersave(dev_priv); intel_uc_fini_misc(dev_priv); diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index 09187286d346..1aaccbe7e1de 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -460,7 +460,7 @@ eb_validate_vma(struct i915_execbuffer *eb, * any non-page-aligned or non-canonical addresses. */ if (unlikely(entry->flags & EXEC_OBJECT_PINNED && - entry->offset != gen8_canonical_addr(entry->offset & PAGE_MASK))) + entry->offset != gen8_canonical_addr(entry->offset & I915_GTT_PAGE_MASK))) return -EINVAL; /* pad_to_size was once a reserved field, so sanitize it */ diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 56c7f8637311..07999fe09ad2 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -1757,7 +1757,7 @@ static void gen6_dump_ppgtt(struct i915_hw_ppgtt *base, struct seq_file *m) if (i == 4) continue; - seq_printf(m, "\t\t(%03d, %04d) %08lx: ", + seq_printf(m, "\t\t(%03d, %04d) %08llx: ", pde, pte, (pde * GEN6_PTES + pte) * I915_GTT_PAGE_SIZE); for (i = 0; i < 4; i++) { @@ -3413,6 +3413,11 @@ static int gen8_gmch_probe(struct i915_ggtt *ggtt) ggtt->vm.insert_page = bxt_vtd_ggtt_insert_page__BKL; if (ggtt->vm.clear_range != nop_clear_range) ggtt->vm.clear_range = bxt_vtd_ggtt_clear_range__BKL; + + /* Prevent recursively calling stop_machine() and deadlocks. */ + dev_info(dev_priv->drm.dev, + "Disabling error capture for VT-d workaround\n"); + i915_disable_error_state(dev_priv, -ENODEV); } ggtt->invalidate = gen6_ggtt_invalidate; diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index 7e2af5f4f39b..28039290655c 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -42,13 +42,15 @@ #include "i915_selftest.h" #include "i915_timeline.h" -#define I915_GTT_PAGE_SIZE_4K BIT(12) -#define I915_GTT_PAGE_SIZE_64K BIT(16) -#define I915_GTT_PAGE_SIZE_2M BIT(21) +#define I915_GTT_PAGE_SIZE_4K BIT_ULL(12) +#define I915_GTT_PAGE_SIZE_64K BIT_ULL(16) +#define I915_GTT_PAGE_SIZE_2M BIT_ULL(21) #define I915_GTT_PAGE_SIZE I915_GTT_PAGE_SIZE_4K #define I915_GTT_MAX_PAGE_SIZE I915_GTT_PAGE_SIZE_2M +#define I915_GTT_PAGE_MASK -I915_GTT_PAGE_SIZE + #define I915_GTT_MIN_ALIGNMENT I915_GTT_PAGE_SIZE #define I915_FENCE_REG_NONE -1 @@ -659,20 +661,20 @@ int i915_gem_gtt_insert(struct i915_address_space *vm, u64 start, u64 end, unsigned int flags); /* Flags used by pin/bind&friends. */ -#define PIN_NONBLOCK BIT(0) -#define PIN_MAPPABLE BIT(1) -#define PIN_ZONE_4G BIT(2) -#define PIN_NONFAULT BIT(3) -#define PIN_NOEVICT BIT(4) - -#define PIN_MBZ BIT(5) /* I915_VMA_PIN_OVERFLOW */ -#define PIN_GLOBAL BIT(6) /* I915_VMA_GLOBAL_BIND */ -#define PIN_USER BIT(7) /* I915_VMA_LOCAL_BIND */ -#define PIN_UPDATE BIT(8) - -#define PIN_HIGH BIT(9) -#define PIN_OFFSET_BIAS BIT(10) -#define PIN_OFFSET_FIXED BIT(11) +#define PIN_NONBLOCK BIT_ULL(0) +#define PIN_MAPPABLE BIT_ULL(1) +#define PIN_ZONE_4G BIT_ULL(2) +#define PIN_NONFAULT BIT_ULL(3) +#define PIN_NOEVICT BIT_ULL(4) + +#define PIN_MBZ BIT_ULL(5) /* I915_VMA_PIN_OVERFLOW */ +#define PIN_GLOBAL BIT_ULL(6) /* I915_VMA_GLOBAL_BIND */ +#define PIN_USER BIT_ULL(7) /* I915_VMA_LOCAL_BIND */ +#define PIN_UPDATE BIT_ULL(8) + +#define PIN_HIGH BIT_ULL(9) +#define PIN_OFFSET_BIAS BIT_ULL(10) +#define PIN_OFFSET_FIXED BIT_ULL(11) #define PIN_OFFSET_MASK (-I915_GTT_PAGE_SIZE) #endif diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index 8762d17b6659..db4128d6c09b 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -648,6 +648,9 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, return 0; } + if (IS_ERR(error)) + return PTR_ERR(error); + if (*error->error_msg) err_printf(m, "%s\n", error->error_msg); err_printf(m, "Kernel: " UTS_RELEASE "\n"); @@ -1492,7 +1495,7 @@ static void gem_record_rings(struct i915_gpu_state *error) if (HAS_BROKEN_CS_TLB(i915)) ee->wa_batchbuffer = i915_error_object_create(i915, - engine->scratch); + i915->gt.scratch); request_record_user_bo(request, ee); ee->ctx = @@ -1859,6 +1862,7 @@ void i915_capture_error_state(struct drm_i915_private *i915, error = i915_capture_gpu_state(i915); if (!error) { DRM_DEBUG_DRIVER("out of memory, not capturing error state\n"); + i915_disable_error_state(i915, -ENOMEM); return; } @@ -1914,5 +1918,14 @@ void i915_reset_error_state(struct drm_i915_private *i915) i915->gpu_error.first_error = NULL; spin_unlock_irq(&i915->gpu_error.lock); - i915_gpu_state_put(error); + if (!IS_ERR(error)) + i915_gpu_state_put(error); +} + +void i915_disable_error_state(struct drm_i915_private *i915, int err) +{ + spin_lock_irq(&i915->gpu_error.lock); + if (!i915->gpu_error.first_error) + i915->gpu_error.first_error = ERR_PTR(err); + spin_unlock_irq(&i915->gpu_error.lock); } diff --git a/drivers/gpu/drm/i915/i915_gpu_error.h b/drivers/gpu/drm/i915/i915_gpu_error.h index 8710fb18ed74..3ec89a504de5 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.h +++ b/drivers/gpu/drm/i915/i915_gpu_error.h @@ -343,6 +343,7 @@ static inline void i915_gpu_state_put(struct i915_gpu_state *gpu) struct i915_gpu_state *i915_first_error_state(struct drm_i915_private *i915); void i915_reset_error_state(struct drm_i915_private *i915); +void i915_disable_error_state(struct drm_i915_private *i915, int err); #else @@ -355,13 +356,18 @@ static inline void i915_capture_error_state(struct drm_i915_private *dev_priv, static inline struct i915_gpu_state * i915_first_error_state(struct drm_i915_private *i915) { - return NULL; + return ERR_PTR(-ENODEV); } static inline void i915_reset_error_state(struct drm_i915_private *i915) { } +static inline void i915_disable_error_state(struct drm_i915_private *i915, + int err) +{ +} + #endif /* IS_ENABLED(CONFIG_DRM_I915_CAPTURE_ERROR) */ #endif /* _I915_GPU_ERROR_H_ */ diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 7c491ea3d052..e31c27e45734 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -2095,8 +2095,12 @@ enum i915_power_well_id { /* ICL PHY DFLEX registers */ #define PORT_TX_DFLEXDPMLE1 _MMIO(0x1638C0) -#define DFLEXDPMLE1_DPMLETC_MASK(n) (0xf << (4 * (n))) -#define DFLEXDPMLE1_DPMLETC(n, x) ((x) << (4 * (n))) +#define DFLEXDPMLE1_DPMLETC_MASK(tc_port) (0xf << (4 * (tc_port))) +#define DFLEXDPMLE1_DPMLETC_ML0(tc_port) (1 << (4 * (tc_port))) +#define DFLEXDPMLE1_DPMLETC_ML1_0(tc_port) (3 << (4 * (tc_port))) +#define DFLEXDPMLE1_DPMLETC_ML3(tc_port) (8 << (4 * (tc_port))) +#define DFLEXDPMLE1_DPMLETC_ML3_2(tc_port) (12 << (4 * (tc_port))) +#define DFLEXDPMLE1_DPMLETC_ML3_0(tc_port) (15 << (4 * (tc_port))) /* BXT PHY Ref registers */ #define _PORT_REF_DW3_A 0x16218C @@ -4593,12 +4597,12 @@ enum { #define DRM_DIP_ENABLE (1 << 28) #define PSR_VSC_BIT_7_SET (1 << 27) -#define VSC_SELECT_MASK (0x3 << 26) -#define VSC_SELECT_SHIFT 26 -#define VSC_DIP_HW_HEA_DATA (0 << 26) -#define VSC_DIP_HW_HEA_SW_DATA (1 << 26) -#define VSC_DIP_HW_DATA_SW_HEA (2 << 26) -#define VSC_DIP_SW_HEA_DATA (3 << 26) +#define VSC_SELECT_MASK (0x3 << 25) +#define VSC_SELECT_SHIFT 25 +#define VSC_DIP_HW_HEA_DATA (0 << 25) +#define VSC_DIP_HW_HEA_SW_DATA (1 << 25) +#define VSC_DIP_HW_DATA_SW_HEA (2 << 25) +#define VSC_DIP_SW_HEA_DATA (3 << 25) #define VDIP_ENABLE_PPS (1 << 24) /* Panel power sequencing */ diff --git a/drivers/gpu/drm/i915/intel_audio.c b/drivers/gpu/drm/i915/intel_audio.c index 769f3f586661..ee3ca2de983b 100644 --- a/drivers/gpu/drm/i915/intel_audio.c +++ b/drivers/gpu/drm/i915/intel_audio.c @@ -144,6 +144,9 @@ static const struct { /* HDMI N/CTS table */ #define TMDS_297M 297000 #define TMDS_296M 296703 +#define TMDS_594M 594000 +#define TMDS_593M 593407 + static const struct { int sample_rate; int clock; @@ -164,6 +167,20 @@ static const struct { { 176400, TMDS_297M, 18816, 247500 }, { 192000, TMDS_296M, 23296, 281250 }, { 192000, TMDS_297M, 20480, 247500 }, + { 44100, TMDS_593M, 8918, 937500 }, + { 44100, TMDS_594M, 9408, 990000 }, + { 48000, TMDS_593M, 5824, 562500 }, + { 48000, TMDS_594M, 6144, 594000 }, + { 32000, TMDS_593M, 5824, 843750 }, + { 32000, TMDS_594M, 3072, 445500 }, + { 88200, TMDS_593M, 17836, 937500 }, + { 88200, TMDS_594M, 18816, 990000 }, + { 96000, TMDS_593M, 11648, 562500 }, + { 96000, TMDS_594M, 12288, 594000 }, + { 176400, TMDS_593M, 35672, 937500 }, + { 176400, TMDS_594M, 37632, 990000 }, + { 192000, TMDS_593M, 23296, 562500 }, + { 192000, TMDS_594M, 24576, 594000 }, }; /* get AUD_CONFIG_PIXEL_CLOCK_HDMI_* value for mode */ diff --git a/drivers/gpu/drm/i915/intel_cdclk.c b/drivers/gpu/drm/i915/intel_cdclk.c index 29075c763428..8d74276029e6 100644 --- a/drivers/gpu/drm/i915/intel_cdclk.c +++ b/drivers/gpu/drm/i915/intel_cdclk.c @@ -2138,16 +2138,8 @@ void intel_set_cdclk(struct drm_i915_private *dev_priv, static int intel_pixel_rate_to_cdclk(struct drm_i915_private *dev_priv, int pixel_rate) { - if (INTEL_GEN(dev_priv) >= 10) + if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) return DIV_ROUND_UP(pixel_rate, 2); - else if (IS_GEMINILAKE(dev_priv)) - /* - * FIXME: Avoid using a pixel clock that is more than 99% of the cdclk - * as a temporary workaround. Use a higher cdclk instead. (Note that - * intel_compute_max_dotclk() limits the max pixel clock to 99% of max - * cdclk.) - */ - return DIV_ROUND_UP(pixel_rate * 100, 2 * 99); else if (IS_GEN9(dev_priv) || IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv)) return pixel_rate; @@ -2543,14 +2535,8 @@ static int intel_compute_max_dotclk(struct drm_i915_private *dev_priv) { int max_cdclk_freq = dev_priv->max_cdclk_freq; - if (INTEL_GEN(dev_priv) >= 10) + if (INTEL_GEN(dev_priv) >= 10 || IS_GEMINILAKE(dev_priv)) return 2 * max_cdclk_freq; - else if (IS_GEMINILAKE(dev_priv)) - /* - * FIXME: Limiting to 99% as a temporary workaround. See - * intel_min_cdclk() for details. - */ - return 2 * max_cdclk_freq * 99 / 100; else if (IS_GEN9(dev_priv) || IS_BROADWELL(dev_priv) || IS_HASWELL(dev_priv)) return max_cdclk_freq; diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c index 0ef0c6448d53..01fa98299bae 100644 --- a/drivers/gpu/drm/i915/intel_device_info.c +++ b/drivers/gpu/drm/i915/intel_device_info.c @@ -474,7 +474,7 @@ static void broadwell_sseu_info_init(struct drm_i915_private *dev_priv) u8 eu_disabled_mask; u32 n_disabled; - if (!(sseu->subslice_mask[ss] & BIT(ss))) + if (!(sseu->subslice_mask[s] & BIT(ss))) /* skip disabled subslice */ continue; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9741cc419e1b..c9878dd1f7cd 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2890,6 +2890,7 @@ intel_find_initial_plane_obj(struct intel_crtc *intel_crtc, return; valid_fb: + intel_state->base.rotation = plane_config->rotation; intel_fill_fb_ggtt_view(&intel_state->view, fb, intel_state->base.rotation); intel_state->color_plane[0].stride = @@ -4850,8 +4851,31 @@ static void cpt_verify_modeset(struct drm_device *dev, int pipe) * chroma samples for both of the luma samples, and thus we don't * actually get the expected MPEG2 chroma siting convention :( * The same behaviour is observed on pre-SKL platforms as well. + * + * Theory behind the formula (note that we ignore sub-pixel + * source coordinates): + * s = source sample position + * d = destination sample position + * + * Downscaling 4:1: + * -0.5 + * | 0.0 + * | | 1.5 (initial phase) + * | | | + * v v v + * | s | s | s | s | + * | d | + * + * Upscaling 1:4: + * -0.5 + * | -0.375 (initial phase) + * | | 0.0 + * | | | + * v v v + * | s | + * | d | d | d | d | */ -u16 skl_scaler_calc_phase(int sub, bool chroma_cosited) +u16 skl_scaler_calc_phase(int sub, int scale, bool chroma_cosited) { int phase = -0x8000; u16 trip = 0; @@ -4859,6 +4883,15 @@ u16 skl_scaler_calc_phase(int sub, bool chroma_cosited) if (chroma_cosited) phase += (sub - 1) * 0x8000 / sub; + phase += scale / (2 * sub); + + /* + * Hardware initial phase limited to [-0.5:1.5]. + * Since the max hardware scale factor is 3.0, we + * should never actually excdeed 1.0 here. + */ + WARN_ON(phase < -0x8000 || phase > 0x18000); + if (phase < 0) phase = 0x10000 + phase; else @@ -5067,13 +5100,20 @@ static void skylake_pfit_enable(struct intel_crtc *crtc) if (crtc->config->pch_pfit.enabled) { u16 uv_rgb_hphase, uv_rgb_vphase; + int pfit_w, pfit_h, hscale, vscale; int id; if (WARN_ON(crtc->config->scaler_state.scaler_id < 0)) return; - uv_rgb_hphase = skl_scaler_calc_phase(1, false); - uv_rgb_vphase = skl_scaler_calc_phase(1, false); + pfit_w = (crtc->config->pch_pfit.size >> 16) & 0xFFFF; + pfit_h = crtc->config->pch_pfit.size & 0xFFFF; + + hscale = (crtc->config->pipe_src_w << 16) / pfit_w; + vscale = (crtc->config->pipe_src_h << 16) / pfit_h; + + uv_rgb_hphase = skl_scaler_calc_phase(1, hscale, false); + uv_rgb_vphase = skl_scaler_calc_phase(1, vscale, false); id = scaler_state->scaler_id; I915_WRITE(SKL_PS_CTRL(pipe, id), PS_SCALER_EN | @@ -7843,8 +7883,15 @@ i9xx_get_initial_plane_config(struct intel_crtc *crtc, plane_config->tiling = I915_TILING_X; fb->modifier = I915_FORMAT_MOD_X_TILED; } + + if (val & DISPPLANE_ROTATE_180) + plane_config->rotation = DRM_MODE_ROTATE_180; } + if (IS_CHERRYVIEW(dev_priv) && pipe == PIPE_B && + val & DISPPLANE_MIRROR) + plane_config->rotation |= DRM_MODE_REFLECT_X; + pixel_format = val & DISPPLANE_PIXFORMAT_MASK; fourcc = i9xx_format_to_fourcc(pixel_format); fb->format = drm_format_info(fourcc); @@ -8913,6 +8960,29 @@ skylake_get_initial_plane_config(struct intel_crtc *crtc, goto error; } + /* + * DRM_MODE_ROTATE_ is counter clockwise to stay compatible with Xrandr + * while i915 HW rotation is clockwise, thats why this swapping. + */ + switch (val & PLANE_CTL_ROTATE_MASK) { + case PLANE_CTL_ROTATE_0: + plane_config->rotation = DRM_MODE_ROTATE_0; + break; + case PLANE_CTL_ROTATE_90: + plane_config->rotation = DRM_MODE_ROTATE_270; + break; + case PLANE_CTL_ROTATE_180: + plane_config->rotation = DRM_MODE_ROTATE_180; + break; + case PLANE_CTL_ROTATE_270: + plane_config->rotation = DRM_MODE_ROTATE_90; + break; + } + + if (INTEL_GEN(dev_priv) >= 10 && + val & PLANE_CTL_FLIP_HORIZONTAL) + plane_config->rotation |= DRM_MODE_REFLECT_X; + base = I915_READ(PLANE_SURF(pipe, plane_id)) & 0xfffff000; plane_config->base = base; @@ -12768,17 +12838,12 @@ static void intel_atomic_commit_tail(struct drm_atomic_state *state) intel_check_cpu_fifo_underruns(dev_priv); intel_check_pch_fifo_underruns(dev_priv); - if (!new_crtc_state->active) { - /* - * Make sure we don't call initial_watermarks - * for ILK-style watermark updates. - * - * No clue what this is supposed to achieve. - */ - if (INTEL_GEN(dev_priv) >= 9) - dev_priv->display.initial_watermarks(intel_state, - to_intel_crtc_state(new_crtc_state)); - } + /* FIXME unify this for all platforms */ + if (!new_crtc_state->active && + !HAS_GMCH_DISPLAY(dev_priv) && + dev_priv->display.initial_watermarks) + dev_priv->display.initial_watermarks(intel_state, + to_intel_crtc_state(new_crtc_state)); } } @@ -14646,7 +14711,7 @@ static int intel_framebuffer_init(struct intel_framebuffer *intel_fb, fb->height < SKL_MIN_YUV_420_SRC_H || (fb->width % 4) != 0 || (fb->height % 4) != 0)) { DRM_DEBUG_KMS("src dimensions not correct for NV12\n"); - return -EINVAL; + goto err; } for (i = 0; i < fb->format->num_planes; i++) { @@ -15233,6 +15298,14 @@ retry: ret = drm_atomic_add_affected_planes(state, crtc); if (ret) goto out; + + /* + * FIXME hack to force a LUT update to avoid the + * plane update forcing the pipe gamma on without + * having a proper LUT loaded. Remove once we + * have readout for pipe gamma enable. + */ + crtc_state->color_mgmt_changed = true; } } diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c index 1b00f8ea145b..a911691dbd0f 100644 --- a/drivers/gpu/drm/i915/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/intel_dp_mst.c @@ -452,6 +452,10 @@ static struct drm_connector *intel_dp_add_mst_connector(struct drm_dp_mst_topolo if (!intel_connector) return NULL; + intel_connector->get_hw_state = intel_dp_mst_get_hw_state; + intel_connector->mst_port = intel_dp; + intel_connector->port = port; + connector = &intel_connector->base; ret = drm_connector_init(dev, connector, &intel_dp_mst_connector_funcs, DRM_MODE_CONNECTOR_DisplayPort); @@ -462,10 +466,6 @@ static struct drm_connector *intel_dp_add_mst_connector(struct drm_dp_mst_topolo drm_connector_helper_add(connector, &intel_dp_mst_connector_helper_funcs); - intel_connector->get_hw_state = intel_dp_mst_get_hw_state; - intel_connector->mst_port = intel_dp; - intel_connector->port = port; - for_each_pipe(dev_priv, pipe) { struct drm_encoder *enc = &intel_dp->mst_encoders[pipe]->base.base; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index f8dc84b2d2d3..db6fa1d0cbda 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -547,6 +547,7 @@ struct intel_initial_plane_config { unsigned int tiling; int size; u32 base; + u8 rotation; }; #define SKL_MIN_SRC_W 8 @@ -1646,7 +1647,7 @@ void intel_mode_from_pipe_config(struct drm_display_mode *mode, void intel_crtc_arm_fifo_underrun(struct intel_crtc *crtc, struct intel_crtc_state *crtc_state); -u16 skl_scaler_calc_phase(int sub, bool chroma_center); +u16 skl_scaler_calc_phase(int sub, int scale, bool chroma_center); int skl_update_scaler_crtc(struct intel_crtc_state *crtc_state); int skl_max_scale(const struct intel_crtc_state *crtc_state, u32 pixel_format); diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c index 217ed3ee1cab..76b5f94ea6cb 100644 --- a/drivers/gpu/drm/i915/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/intel_engine_cs.c @@ -490,46 +490,6 @@ void intel_engine_setup_common(struct intel_engine_cs *engine) intel_engine_init_cmd_parser(engine); } -int intel_engine_create_scratch(struct intel_engine_cs *engine, - unsigned int size) -{ - struct drm_i915_gem_object *obj; - struct i915_vma *vma; - int ret; - - WARN_ON(engine->scratch); - - obj = i915_gem_object_create_stolen(engine->i915, size); - if (!obj) - obj = i915_gem_object_create_internal(engine->i915, size); - if (IS_ERR(obj)) { - DRM_ERROR("Failed to allocate scratch page\n"); - return PTR_ERR(obj); - } - - vma = i915_vma_instance(obj, &engine->i915->ggtt.vm, NULL); - if (IS_ERR(vma)) { - ret = PTR_ERR(vma); - goto err_unref; - } - - ret = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH); - if (ret) - goto err_unref; - - engine->scratch = vma; - return 0; - -err_unref: - i915_gem_object_put(obj); - return ret; -} - -void intel_engine_cleanup_scratch(struct intel_engine_cs *engine) -{ - i915_vma_unpin_and_release(&engine->scratch, 0); -} - static void cleanup_status_page(struct intel_engine_cs *engine) { if (HWS_NEEDS_PHYSICAL(engine->i915)) { @@ -704,8 +664,6 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine) { struct drm_i915_private *i915 = engine->i915; - intel_engine_cleanup_scratch(engine); - cleanup_status_page(engine); intel_engine_fini_breadcrumbs(engine); @@ -720,6 +678,8 @@ void intel_engine_cleanup_common(struct intel_engine_cs *engine) __intel_context_unpin(i915->kernel_context, engine); i915_timeline_fini(&engine->timeline); + + intel_wa_list_free(&engine->wa_list); } u64 intel_engine_get_active_head(const struct intel_engine_cs *engine) diff --git a/drivers/gpu/drm/i915/intel_hotplug.c b/drivers/gpu/drm/i915/intel_hotplug.c index 648a13c6043c..9a8018130237 100644 --- a/drivers/gpu/drm/i915/intel_hotplug.c +++ b/drivers/gpu/drm/i915/intel_hotplug.c @@ -228,7 +228,9 @@ static void intel_hpd_irq_storm_reenable_work(struct work_struct *work) drm_for_each_connector_iter(connector, &conn_iter) { struct intel_connector *intel_connector = to_intel_connector(connector); - if (intel_connector->encoder->hpd_pin == pin) { + /* Don't check MST ports, they don't have pins */ + if (!intel_connector->mst_port && + intel_connector->encoder->hpd_pin == pin) { if (connector->polled != intel_connector->polled) DRM_DEBUG_DRIVER("Reenabling HPD on connector %s\n", connector->name); @@ -395,37 +397,54 @@ void intel_hpd_irq_handler(struct drm_i915_private *dev_priv, struct intel_encoder *encoder; bool storm_detected = false; bool queue_dig = false, queue_hp = false; + u32 long_hpd_pulse_mask = 0; + u32 short_hpd_pulse_mask = 0; + enum hpd_pin pin; if (!pin_mask) return; spin_lock(&dev_priv->irq_lock); + + /* + * Determine whether ->hpd_pulse() exists for each pin, and + * whether we have a short or a long pulse. This is needed + * as each pin may have up to two encoders (HDMI and DP) and + * only the one of them (DP) will have ->hpd_pulse(). + */ for_each_intel_encoder(&dev_priv->drm, encoder) { - enum hpd_pin pin = encoder->hpd_pin; bool has_hpd_pulse = intel_encoder_has_hpd_pulse(encoder); + enum port port = encoder->port; + bool long_hpd; + pin = encoder->hpd_pin; if (!(BIT(pin) & pin_mask)) continue; - if (has_hpd_pulse) { - bool long_hpd = long_mask & BIT(pin); - enum port port = encoder->port; + if (!has_hpd_pulse) + continue; - DRM_DEBUG_DRIVER("digital hpd port %c - %s\n", port_name(port), - long_hpd ? "long" : "short"); - /* - * For long HPD pulses we want to have the digital queue happen, - * but we still want HPD storm detection to function. - */ - queue_dig = true; - if (long_hpd) { - dev_priv->hotplug.long_port_mask |= (1 << port); - } else { - /* for short HPD just trigger the digital queue */ - dev_priv->hotplug.short_port_mask |= (1 << port); - continue; - } + long_hpd = long_mask & BIT(pin); + + DRM_DEBUG_DRIVER("digital hpd port %c - %s\n", port_name(port), + long_hpd ? "long" : "short"); + queue_dig = true; + + if (long_hpd) { + long_hpd_pulse_mask |= BIT(pin); + dev_priv->hotplug.long_port_mask |= BIT(port); + } else { + short_hpd_pulse_mask |= BIT(pin); + dev_priv->hotplug.short_port_mask |= BIT(port); } + } + + /* Now process each pin just once */ + for_each_hpd_pin(pin) { + bool long_hpd; + + if (!(BIT(pin) & pin_mask)) + continue; if (dev_priv->hotplug.stats[pin].state == HPD_DISABLED) { /* @@ -442,11 +461,22 @@ void intel_hpd_irq_handler(struct drm_i915_private *dev_priv, if (dev_priv->hotplug.stats[pin].state != HPD_ENABLED) continue; - if (!has_hpd_pulse) { + /* + * Delegate to ->hpd_pulse() if one of the encoders for this + * pin has it, otherwise let the hotplug_work deal with this + * pin directly. + */ + if (((short_hpd_pulse_mask | long_hpd_pulse_mask) & BIT(pin))) { + long_hpd = long_hpd_pulse_mask & BIT(pin); + } else { dev_priv->hotplug.event_bits |= BIT(pin); + long_hpd = true; queue_hp = true; } + if (!long_hpd) + continue; + if (intel_hpd_irq_storm_detect(dev_priv, pin)) { dev_priv->hotplug.event_bits &= ~BIT(pin); storm_detected = true; diff --git a/drivers/gpu/drm/i915/intel_lpe_audio.c b/drivers/gpu/drm/i915/intel_lpe_audio.c index cdf19553ffac..5d5336fbe7b0 100644 --- a/drivers/gpu/drm/i915/intel_lpe_audio.c +++ b/drivers/gpu/drm/i915/intel_lpe_audio.c @@ -297,8 +297,10 @@ void intel_lpe_audio_teardown(struct drm_i915_private *dev_priv) lpe_audio_platdev_destroy(dev_priv); irq_free_desc(dev_priv->lpe_audio.irq); -} + dev_priv->lpe_audio.irq = -1; + dev_priv->lpe_audio.platdev = NULL; +} /** * intel_lpe_audio_notify() - notify lpe audio event diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 43957bb37a42..58d1d3d47dd3 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -424,7 +424,8 @@ static u64 execlists_update_context(struct i915_request *rq) reg_state[CTX_RING_TAIL+1] = intel_ring_set_tail(rq->ring, rq->tail); - /* True 32b PPGTT with dynamic page allocation: update PDP + /* + * True 32b PPGTT with dynamic page allocation: update PDP * registers and point the unallocated PDPs to scratch page. * PML4 is allocated during ppgtt init, so this is not needed * in 48-bit mode. @@ -432,6 +433,22 @@ static u64 execlists_update_context(struct i915_request *rq) if (ppgtt && !i915_vm_is_48bit(&ppgtt->vm)) execlists_update_context_pdps(ppgtt, reg_state); + /* + * Make sure the context image is complete before we submit it to HW. + * + * Ostensibly, writes (including the WCB) should be flushed prior to + * an uncached write such as our mmio register access, the empirical + * evidence (esp. on Braswell) suggests that the WC write into memory + * may not be visible to the HW prior to the completion of the UC + * register write and that we may begin execution from the context + * before its image is complete leading to invalid PD chasing. + * + * Furthermore, Braswell, at least, wants a full mb to be sure that + * the writes are coherent in memory (visible to the GPU) prior to + * execution, and not just visible to other CPUs (as is the result of + * wmb). + */ + mb(); return ce->lrc_desc; } @@ -1431,9 +1448,10 @@ static int execlists_request_alloc(struct i915_request *request) static u32 * gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch) { + /* NB no one else is allowed to scribble over scratch + 256! */ *batch++ = MI_STORE_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT; *batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4); - *batch++ = i915_ggtt_offset(engine->scratch) + 256; + *batch++ = i915_scratch_offset(engine->i915) + 256; *batch++ = 0; *batch++ = MI_LOAD_REGISTER_IMM(1); @@ -1447,7 +1465,7 @@ gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch) *batch++ = MI_LOAD_REGISTER_MEM_GEN8 | MI_SRM_LRM_GLOBAL_GTT; *batch++ = i915_mmio_reg_offset(GEN8_L3SQCREG4); - *batch++ = i915_ggtt_offset(engine->scratch) + 256; + *batch++ = i915_scratch_offset(engine->i915) + 256; *batch++ = 0; return batch; @@ -1484,7 +1502,7 @@ static u32 *gen8_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch) PIPE_CONTROL_GLOBAL_GTT_IVB | PIPE_CONTROL_CS_STALL | PIPE_CONTROL_QW_WRITE, - i915_ggtt_offset(engine->scratch) + + i915_scratch_offset(engine->i915) + 2 * CACHELINE_BYTES); *batch++ = MI_ARB_ON_OFF | MI_ARB_ENABLE; @@ -1561,7 +1579,7 @@ static u32 *gen9_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch) PIPE_CONTROL_GLOBAL_GTT_IVB | PIPE_CONTROL_CS_STALL | PIPE_CONTROL_QW_WRITE, - i915_ggtt_offset(engine->scratch) + i915_scratch_offset(engine->i915) + 2 * CACHELINE_BYTES); } @@ -1781,6 +1799,8 @@ static bool unexpected_starting_state(struct intel_engine_cs *engine) static int gen8_init_common_ring(struct intel_engine_cs *engine) { + intel_engine_apply_workarounds(engine); + intel_mocs_init_engine(engine); intel_engine_reset_breadcrumbs(engine); @@ -2127,7 +2147,7 @@ static int gen8_emit_flush_render(struct i915_request *request, { struct intel_engine_cs *engine = request->engine; u32 scratch_addr = - i915_ggtt_offset(engine->scratch) + 2 * CACHELINE_BYTES; + i915_scratch_offset(engine->i915) + 2 * CACHELINE_BYTES; bool vf_flush_wa = false, dc_flush_wa = false; u32 *cs, flags = 0; int len; @@ -2464,10 +2484,6 @@ int logical_render_ring_init(struct intel_engine_cs *engine) if (ret) return ret; - ret = intel_engine_create_scratch(engine, PAGE_SIZE); - if (ret) - goto err_cleanup_common; - ret = intel_init_workaround_bb(engine); if (ret) { /* @@ -2479,11 +2495,9 @@ int logical_render_ring_init(struct intel_engine_cs *engine) ret); } - return 0; + intel_engine_init_workarounds(engine); -err_cleanup_common: - intel_engine_cleanup_common(engine); - return ret; + return 0; } int logical_xcs_ring_init(struct intel_engine_cs *engine) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 1db9b8328275..3fe358db1276 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2493,6 +2493,9 @@ static uint32_t ilk_compute_pri_wm(const struct intel_crtc_state *cstate, uint32_t method1, method2; int cpp; + if (mem_value == 0) + return U32_MAX; + if (!intel_wm_plane_visible(cstate, pstate)) return 0; @@ -2522,6 +2525,9 @@ static uint32_t ilk_compute_spr_wm(const struct intel_crtc_state *cstate, uint32_t method1, method2; int cpp; + if (mem_value == 0) + return U32_MAX; + if (!intel_wm_plane_visible(cstate, pstate)) return 0; @@ -2545,6 +2551,9 @@ static uint32_t ilk_compute_cur_wm(const struct intel_crtc_state *cstate, { int cpp; + if (mem_value == 0) + return U32_MAX; + if (!intel_wm_plane_visible(cstate, pstate)) return 0; @@ -2881,8 +2890,7 @@ static void intel_read_wm_latency(struct drm_i915_private *dev_priv, * any underrun. If not able to get Dimm info assume 16GB dimm * to avoid any underrun. */ - if (!dev_priv->dram_info.valid_dimm || - dev_priv->dram_info.is_16gb_dimm) + if (dev_priv->dram_info.is_16gb_dimm) wm[0] += 1; } else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { @@ -3009,6 +3017,34 @@ static void snb_wm_latency_quirk(struct drm_i915_private *dev_priv) intel_print_wm_latency(dev_priv, "Cursor", dev_priv->wm.cur_latency); } +static void snb_wm_lp3_irq_quirk(struct drm_i915_private *dev_priv) +{ + /* + * On some SNB machines (Thinkpad X220 Tablet at least) + * LP3 usage can cause vblank interrupts to be lost. + * The DEIIR bit will go high but it looks like the CPU + * never gets interrupted. + * + * It's not clear whether other interrupt source could + * be affected or if this is somehow limited to vblank + * interrupts only. To play it safe we disable LP3 + * watermarks entirely. + */ + if (dev_priv->wm.pri_latency[3] == 0 && + dev_priv->wm.spr_latency[3] == 0 && + dev_priv->wm.cur_latency[3] == 0) + return; + + dev_priv->wm.pri_latency[3] = 0; + dev_priv->wm.spr_latency[3] = 0; + dev_priv->wm.cur_latency[3] = 0; + + DRM_DEBUG_KMS("LP3 watermarks disabled due to potential for lost interrupts\n"); + intel_print_wm_latency(dev_priv, "Primary", dev_priv->wm.pri_latency); + intel_print_wm_latency(dev_priv, "Sprite", dev_priv->wm.spr_latency); + intel_print_wm_latency(dev_priv, "Cursor", dev_priv->wm.cur_latency); +} + static void ilk_setup_wm_latency(struct drm_i915_private *dev_priv) { intel_read_wm_latency(dev_priv, dev_priv->wm.pri_latency); @@ -3025,8 +3061,10 @@ static void ilk_setup_wm_latency(struct drm_i915_private *dev_priv) intel_print_wm_latency(dev_priv, "Sprite", dev_priv->wm.spr_latency); intel_print_wm_latency(dev_priv, "Cursor", dev_priv->wm.cur_latency); - if (IS_GEN6(dev_priv)) + if (IS_GEN6(dev_priv)) { snb_wm_latency_quirk(dev_priv); + snb_wm_lp3_irq_quirk(dev_priv); + } } static void skl_setup_wm_latency(struct drm_i915_private *dev_priv) diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index d0ef50bf930a..1f8d2a66c791 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -69,19 +69,28 @@ unsigned int intel_ring_update_space(struct intel_ring *ring) static int gen2_render_ring_flush(struct i915_request *rq, u32 mode) { + unsigned int num_store_dw; u32 cmd, *cs; cmd = MI_FLUSH; - + num_store_dw = 0; if (mode & EMIT_INVALIDATE) cmd |= MI_READ_FLUSH; + if (mode & EMIT_FLUSH) + num_store_dw = 4; - cs = intel_ring_begin(rq, 2); + cs = intel_ring_begin(rq, 2 + 3 * num_store_dw); if (IS_ERR(cs)) return PTR_ERR(cs); *cs++ = cmd; - *cs++ = MI_NOOP; + while (num_store_dw--) { + *cs++ = MI_STORE_DWORD_IMM | MI_MEM_VIRTUAL; + *cs++ = i915_scratch_offset(rq->i915); + *cs++ = 0; + } + *cs++ = MI_FLUSH | MI_NO_WRITE_FLUSH; + intel_ring_advance(rq, cs); return 0; @@ -91,6 +100,7 @@ static int gen4_render_ring_flush(struct i915_request *rq, u32 mode) { u32 cmd, *cs; + int i; /* * read/write caches: @@ -127,12 +137,43 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode) cmd |= MI_INVALIDATE_ISP; } - cs = intel_ring_begin(rq, 2); + i = 2; + if (mode & EMIT_INVALIDATE) + i += 20; + + cs = intel_ring_begin(rq, i); if (IS_ERR(cs)) return PTR_ERR(cs); *cs++ = cmd; - *cs++ = MI_NOOP; + + /* + * A random delay to let the CS invalidate take effect? Without this + * delay, the GPU relocation path fails as the CS does not see + * the updated contents. Just as important, if we apply the flushes + * to the EMIT_FLUSH branch (i.e. immediately after the relocation + * write and before the invalidate on the next batch), the relocations + * still fail. This implies that is a delay following invalidation + * that is required to reset the caches as opposed to a delay to + * ensure the memory is written. + */ + if (mode & EMIT_INVALIDATE) { + *cs++ = GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE; + *cs++ = i915_scratch_offset(rq->i915) | PIPE_CONTROL_GLOBAL_GTT; + *cs++ = 0; + *cs++ = 0; + + for (i = 0; i < 12; i++) + *cs++ = MI_FLUSH; + + *cs++ = GFX_OP_PIPE_CONTROL(4) | PIPE_CONTROL_QW_WRITE; + *cs++ = i915_scratch_offset(rq->i915) | PIPE_CONTROL_GLOBAL_GTT; + *cs++ = 0; + *cs++ = 0; + } + + *cs++ = cmd; + intel_ring_advance(rq, cs); return 0; @@ -178,8 +219,7 @@ gen4_render_ring_flush(struct i915_request *rq, u32 mode) static int intel_emit_post_sync_nonzero_flush(struct i915_request *rq) { - u32 scratch_addr = - i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES; + u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES; u32 *cs; cs = intel_ring_begin(rq, 6); @@ -212,8 +252,7 @@ intel_emit_post_sync_nonzero_flush(struct i915_request *rq) static int gen6_render_ring_flush(struct i915_request *rq, u32 mode) { - u32 scratch_addr = - i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES; + u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES; u32 *cs, flags = 0; int ret; @@ -282,8 +321,7 @@ gen7_render_ring_cs_stall_wa(struct i915_request *rq) static int gen7_render_ring_flush(struct i915_request *rq, u32 mode) { - u32 scratch_addr = - i915_ggtt_offset(rq->engine->scratch) + 2 * CACHELINE_BYTES; + u32 scratch_addr = i915_scratch_offset(rq->i915) + 2 * CACHELINE_BYTES; u32 *cs, flags = 0; /* @@ -937,7 +975,7 @@ i965_emit_bb_start(struct i915_request *rq, } /* Just userspace ABI convention to limit the wa batch bo to a resonable size */ -#define I830_BATCH_LIMIT (256*1024) +#define I830_BATCH_LIMIT SZ_256K #define I830_TLB_ENTRIES (2) #define I830_WA_SIZE max(I830_TLB_ENTRIES*4096, I830_BATCH_LIMIT) static int @@ -945,7 +983,9 @@ i830_emit_bb_start(struct i915_request *rq, u64 offset, u32 len, unsigned int dispatch_flags) { - u32 *cs, cs_offset = i915_ggtt_offset(rq->engine->scratch); + u32 *cs, cs_offset = i915_scratch_offset(rq->i915); + + GEM_BUG_ON(rq->i915->gt.scratch->size < I830_WA_SIZE); cs = intel_ring_begin(rq, 6); if (IS_ERR(cs)) @@ -1403,7 +1443,6 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine) { struct i915_timeline *timeline; struct intel_ring *ring; - unsigned int size; int err; intel_engine_setup_common(engine); @@ -1428,21 +1467,12 @@ static int intel_init_ring_buffer(struct intel_engine_cs *engine) GEM_BUG_ON(engine->buffer); engine->buffer = ring; - size = PAGE_SIZE; - if (HAS_BROKEN_CS_TLB(engine->i915)) - size = I830_WA_SIZE; - err = intel_engine_create_scratch(engine, size); - if (err) - goto err_unpin; - err = intel_engine_init_common(engine); if (err) - goto err_scratch; + goto err_unpin; return 0; -err_scratch: - intel_engine_cleanup_scratch(engine); err_unpin: intel_ring_unpin(ring); err_ring: @@ -1516,7 +1546,7 @@ static int flush_pd_dir(struct i915_request *rq) /* Stall until the page table load is complete */ *cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT; *cs++ = i915_mmio_reg_offset(RING_PP_DIR_BASE(engine)); - *cs++ = i915_ggtt_offset(engine->scratch); + *cs++ = i915_scratch_offset(rq->i915); *cs++ = MI_NOOP; intel_ring_advance(rq, cs); @@ -1625,7 +1655,7 @@ static inline int mi_set_context(struct i915_request *rq, u32 flags) /* Insert a delay before the next switch! */ *cs++ = MI_STORE_REGISTER_MEM | MI_SRM_LRM_GLOBAL_GTT; *cs++ = i915_mmio_reg_offset(last_reg); - *cs++ = i915_ggtt_offset(engine->scratch); + *cs++ = i915_scratch_offset(rq->i915); *cs++ = MI_NOOP; } *cs++ = MI_ARB_ON_OFF | MI_ARB_ENABLE; diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 2dfa585712c2..767a7192c969 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -15,6 +15,7 @@ #include "i915_selftest.h" #include "i915_timeline.h" #include "intel_gpu_commands.h" +#include "intel_workarounds.h" struct drm_printer; struct i915_sched_attr; @@ -440,7 +441,7 @@ struct intel_engine_cs { struct intel_hw_status_page status_page; struct i915_ctx_workarounds wa_ctx; - struct i915_vma *scratch; + struct i915_wa_list wa_list; u32 irq_keep_mask; /* always keep these interrupts */ u32 irq_enable_mask; /* bitmask to enable ring interrupt */ @@ -898,10 +899,6 @@ void intel_engine_setup_common(struct intel_engine_cs *engine); int intel_engine_init_common(struct intel_engine_cs *engine); void intel_engine_cleanup_common(struct intel_engine_cs *engine); -int intel_engine_create_scratch(struct intel_engine_cs *engine, - unsigned int size); -void intel_engine_cleanup_scratch(struct intel_engine_cs *engine); - int intel_init_render_ring_buffer(struct intel_engine_cs *engine); int intel_init_bsd_ring_buffer(struct intel_engine_cs *engine); int intel_init_blt_ring_buffer(struct intel_engine_cs *engine); diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c index 0fdabce647ab..44e4491a4918 100644 --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -2749,6 +2749,12 @@ static const struct i915_power_well_desc icl_power_wells[] = { }, }, { + .name = "DC off", + .domains = ICL_DISPLAY_DC_OFF_POWER_DOMAINS, + .ops = &gen9_dc_off_power_well_ops, + .id = DISP_PW_ID_NONE, + }, + { .name = "power well 2", .domains = ICL_PW_2_POWER_DOMAINS, .ops = &hsw_power_well_ops, @@ -2760,12 +2766,6 @@ static const struct i915_power_well_desc icl_power_wells[] = { }, }, { - .name = "DC off", - .domains = ICL_DISPLAY_DC_OFF_POWER_DOMAINS, - .ops = &gen9_dc_off_power_well_ops, - .id = DISP_PW_ID_NONE, - }, - { .name = "power well 3", .domains = ICL_PW_3_POWER_DOMAINS, .ops = &hsw_power_well_ops, @@ -3176,8 +3176,7 @@ static u8 intel_dbuf_max_slices(struct drm_i915_private *dev_priv) void icl_dbuf_slices_update(struct drm_i915_private *dev_priv, u8 req_slices) { - u8 hw_enabled_slices = dev_priv->wm.skl_hw.ddb.enabled_slices; - u32 val; + const u8 hw_enabled_slices = dev_priv->wm.skl_hw.ddb.enabled_slices; bool ret; if (req_slices > intel_dbuf_max_slices(dev_priv)) { @@ -3188,7 +3187,6 @@ void icl_dbuf_slices_update(struct drm_i915_private *dev_priv, if (req_slices == hw_enabled_slices || req_slices == 0) return; - val = I915_READ(DBUF_CTL_S2); if (req_slices > hw_enabled_slices) ret = intel_dbuf_slice_set(dev_priv, DBUF_CTL_S2, true); else diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 5fd2f7bf3927..d3090a7537bb 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -302,13 +302,65 @@ skl_plane_max_stride(struct intel_plane *plane, return min(8192 * cpp, 32768); } +static void +skl_program_scaler(struct intel_plane *plane, + const struct intel_crtc_state *crtc_state, + const struct intel_plane_state *plane_state) +{ + struct drm_i915_private *dev_priv = to_i915(plane->base.dev); + enum pipe pipe = plane->pipe; + int scaler_id = plane_state->scaler_id; + const struct intel_scaler *scaler = + &crtc_state->scaler_state.scalers[scaler_id]; + int crtc_x = plane_state->base.dst.x1; + int crtc_y = plane_state->base.dst.y1; + uint32_t crtc_w = drm_rect_width(&plane_state->base.dst); + uint32_t crtc_h = drm_rect_height(&plane_state->base.dst); + u16 y_hphase, uv_rgb_hphase; + u16 y_vphase, uv_rgb_vphase; + int hscale, vscale; + + hscale = drm_rect_calc_hscale(&plane_state->base.src, + &plane_state->base.dst, + 0, INT_MAX); + vscale = drm_rect_calc_vscale(&plane_state->base.src, + &plane_state->base.dst, + 0, INT_MAX); + + /* TODO: handle sub-pixel coordinates */ + if (plane_state->base.fb->format->format == DRM_FORMAT_NV12) { + y_hphase = skl_scaler_calc_phase(1, hscale, false); + y_vphase = skl_scaler_calc_phase(1, vscale, false); + + /* MPEG2 chroma siting convention */ + uv_rgb_hphase = skl_scaler_calc_phase(2, hscale, true); + uv_rgb_vphase = skl_scaler_calc_phase(2, vscale, false); + } else { + /* not used */ + y_hphase = 0; + y_vphase = 0; + + uv_rgb_hphase = skl_scaler_calc_phase(1, hscale, false); + uv_rgb_vphase = skl_scaler_calc_phase(1, vscale, false); + } + + I915_WRITE_FW(SKL_PS_CTRL(pipe, scaler_id), + PS_SCALER_EN | PS_PLANE_SEL(plane->id) | scaler->mode); + I915_WRITE_FW(SKL_PS_PWR_GATE(pipe, scaler_id), 0); + I915_WRITE_FW(SKL_PS_VPHASE(pipe, scaler_id), + PS_Y_PHASE(y_vphase) | PS_UV_RGB_PHASE(uv_rgb_vphase)); + I915_WRITE_FW(SKL_PS_HPHASE(pipe, scaler_id), + PS_Y_PHASE(y_hphase) | PS_UV_RGB_PHASE(uv_rgb_hphase)); + I915_WRITE_FW(SKL_PS_WIN_POS(pipe, scaler_id), (crtc_x << 16) | crtc_y); + I915_WRITE_FW(SKL_PS_WIN_SZ(pipe, scaler_id), (crtc_w << 16) | crtc_h); +} + void skl_update_plane(struct intel_plane *plane, const struct intel_crtc_state *crtc_state, const struct intel_plane_state *plane_state) { struct drm_i915_private *dev_priv = to_i915(plane->base.dev); - const struct drm_framebuffer *fb = plane_state->base.fb; enum plane_id plane_id = plane->id; enum pipe pipe = plane->pipe; u32 plane_ctl = plane_state->ctl; @@ -318,8 +370,6 @@ skl_update_plane(struct intel_plane *plane, u32 aux_stride = skl_plane_stride(plane_state, 1); int crtc_x = plane_state->base.dst.x1; int crtc_y = plane_state->base.dst.y1; - uint32_t crtc_w = drm_rect_width(&plane_state->base.dst); - uint32_t crtc_h = drm_rect_height(&plane_state->base.dst); uint32_t x = plane_state->color_plane[0].x; uint32_t y = plane_state->color_plane[0].y; uint32_t src_w = drm_rect_width(&plane_state->base.src) >> 16; @@ -329,8 +379,6 @@ skl_update_plane(struct intel_plane *plane, /* Sizes are 0 based */ src_w--; src_h--; - crtc_w--; - crtc_h--; spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); @@ -353,41 +401,8 @@ skl_update_plane(struct intel_plane *plane, (plane_state->color_plane[1].y << 16) | plane_state->color_plane[1].x); - /* program plane scaler */ if (plane_state->scaler_id >= 0) { - int scaler_id = plane_state->scaler_id; - const struct intel_scaler *scaler = - &crtc_state->scaler_state.scalers[scaler_id]; - u16 y_hphase, uv_rgb_hphase; - u16 y_vphase, uv_rgb_vphase; - - /* TODO: handle sub-pixel coordinates */ - if (fb->format->format == DRM_FORMAT_NV12) { - y_hphase = skl_scaler_calc_phase(1, false); - y_vphase = skl_scaler_calc_phase(1, false); - - /* MPEG2 chroma siting convention */ - uv_rgb_hphase = skl_scaler_calc_phase(2, true); - uv_rgb_vphase = skl_scaler_calc_phase(2, false); - } else { - /* not used */ - y_hphase = 0; - y_vphase = 0; - - uv_rgb_hphase = skl_scaler_calc_phase(1, false); - uv_rgb_vphase = skl_scaler_calc_phase(1, false); - } - - I915_WRITE_FW(SKL_PS_CTRL(pipe, scaler_id), - PS_SCALER_EN | PS_PLANE_SEL(plane_id) | scaler->mode); - I915_WRITE_FW(SKL_PS_PWR_GATE(pipe, scaler_id), 0); - I915_WRITE_FW(SKL_PS_VPHASE(pipe, scaler_id), - PS_Y_PHASE(y_vphase) | PS_UV_RGB_PHASE(uv_rgb_vphase)); - I915_WRITE_FW(SKL_PS_HPHASE(pipe, scaler_id), - PS_Y_PHASE(y_hphase) | PS_UV_RGB_PHASE(uv_rgb_hphase)); - I915_WRITE_FW(SKL_PS_WIN_POS(pipe, scaler_id), (crtc_x << 16) | crtc_y); - I915_WRITE_FW(SKL_PS_WIN_SZ(pipe, scaler_id), - ((crtc_w + 1) << 16)|(crtc_h + 1)); + skl_program_scaler(plane, crtc_state, plane_state); I915_WRITE_FW(PLANE_POS(pipe, plane_id), 0); } else { diff --git a/drivers/gpu/drm/i915/intel_workarounds.c b/drivers/gpu/drm/i915/intel_workarounds.c index 4bcdeaf8d98f..6e580891db96 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.c +++ b/drivers/gpu/drm/i915/intel_workarounds.c @@ -48,6 +48,20 @@ * - Public functions to init or apply the given workaround type. */ +static void wa_init_start(struct i915_wa_list *wal, const char *name) +{ + wal->name = name; +} + +static void wa_init_finish(struct i915_wa_list *wal) +{ + if (!wal->count) + return; + + DRM_DEBUG_DRIVER("Initialized %u %s workarounds\n", + wal->count, wal->name); +} + static void wa_add(struct drm_i915_private *i915, i915_reg_t reg, const u32 mask, const u32 val) { @@ -580,160 +594,175 @@ int intel_ctx_workarounds_emit(struct i915_request *rq) return 0; } -static void bdw_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void +wal_add(struct i915_wa_list *wal, const struct i915_wa *wa) +{ + const unsigned int grow = 1 << 4; + + GEM_BUG_ON(!is_power_of_2(grow)); + + if (IS_ALIGNED(wal->count, grow)) { /* Either uninitialized or full. */ + struct i915_wa *list; + + list = kmalloc_array(ALIGN(wal->count + 1, grow), sizeof(*wa), + GFP_KERNEL); + if (!list) { + DRM_ERROR("No space for workaround init!\n"); + return; + } + + if (wal->list) + memcpy(list, wal->list, sizeof(*wa) * wal->count); + + wal->list = list; + } + + wal->list[wal->count++] = *wa; +} + +static void +wa_masked_en(struct i915_wa_list *wal, i915_reg_t reg, u32 val) +{ + struct i915_wa wa = { + .reg = reg, + .mask = val, + .val = _MASKED_BIT_ENABLE(val) + }; + + wal_add(wal, &wa); +} + +static void +wa_write_masked_or(struct i915_wa_list *wal, i915_reg_t reg, u32 mask, + u32 val) { + struct i915_wa wa = { + .reg = reg, + .mask = mask, + .val = val + }; + + wal_add(wal, &wa); } -static void chv_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void +wa_write(struct i915_wa_list *wal, i915_reg_t reg, u32 val) { + wa_write_masked_or(wal, reg, ~0, val); } -static void gen9_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void +wa_write_or(struct i915_wa_list *wal, i915_reg_t reg, u32 val) { - /* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */ - I915_WRITE(GEN9_CSFE_CHICKEN1_RCS, - _MASKED_BIT_ENABLE(GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE)); + wa_write_masked_or(wal, reg, val, val); +} - /* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */ - I915_WRITE(BDW_SCRATCH1, I915_READ(BDW_SCRATCH1) | - GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE); +static void gen9_gt_workarounds_init(struct drm_i915_private *i915) +{ + struct i915_wa_list *wal = &i915->gt_wa_list; /* WaDisableKillLogic:bxt,skl,kbl */ - if (!IS_COFFEELAKE(dev_priv)) - I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | - ECOCHK_DIS_TLB); + if (!IS_COFFEELAKE(i915)) + wa_write_or(wal, + GAM_ECOCHK, + ECOCHK_DIS_TLB); - if (HAS_LLC(dev_priv)) { + if (HAS_LLC(i915)) { /* WaCompressedResourceSamplerPbeMediaNewHashMode:skl,kbl * * Must match Display Engine. See * WaCompressedResourceDisplayNewHashMode. */ - I915_WRITE(MMCD_MISC_CTRL, - I915_READ(MMCD_MISC_CTRL) | - MMCD_PCLA | - MMCD_HOTSPOT_EN); + wa_write_or(wal, + MMCD_MISC_CTRL, + MMCD_PCLA | MMCD_HOTSPOT_EN); } /* WaDisableHDCInvalidation:skl,bxt,kbl,cfl */ - I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | - BDW_DISABLE_HDC_INVALIDATION); - - /* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */ - if (IS_GEN9_LP(dev_priv)) { - u32 val = I915_READ(GEN8_L3SQCREG1); - - val &= ~L3_PRIO_CREDITS_MASK; - val |= L3_GENERAL_PRIO_CREDITS(62) | L3_HIGH_PRIO_CREDITS(2); - I915_WRITE(GEN8_L3SQCREG1, val); - } - - /* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */ - I915_WRITE(GEN8_L3SQCREG4, - I915_READ(GEN8_L3SQCREG4) | GEN8_LQSC_FLUSH_COHERENT_LINES); - - /* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,[cnl] */ - I915_WRITE(GEN7_FF_SLICE_CS_CHICKEN1, - _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL)); + wa_write_or(wal, + GAM_ECOCHK, + BDW_DISABLE_HDC_INVALIDATION); } -static void skl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void skl_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; - /* WaEnableGapsTsvCreditFix:skl */ - I915_WRITE(GEN8_GARBCNTL, - I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE); + gen9_gt_workarounds_init(i915); /* WaDisableGafsUnitClkGating:skl */ - I915_WRITE(GEN7_UCGCTL4, - I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); + wa_write_or(wal, + GEN7_UCGCTL4, + GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); /* WaInPlaceDecompressionHang:skl */ - if (IS_SKL_REVID(dev_priv, SKL_REVID_H0, REVID_FOREVER)) - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + if (IS_SKL_REVID(i915, SKL_REVID_H0, REVID_FOREVER)) + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); } -static void bxt_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void bxt_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; - /* WaDisablePooledEuLoadBalancingFix:bxt */ - I915_WRITE(FF_SLICE_CS_CHICKEN2, - _MASKED_BIT_ENABLE(GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE)); + gen9_gt_workarounds_init(i915); /* WaInPlaceDecompressionHang:bxt */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); } -static void kbl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void kbl_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; - /* WaEnableGapsTsvCreditFix:kbl */ - I915_WRITE(GEN8_GARBCNTL, - I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE); + gen9_gt_workarounds_init(i915); /* WaDisableDynamicCreditSharing:kbl */ - if (IS_KBL_REVID(dev_priv, 0, KBL_REVID_B0)) - I915_WRITE(GAMT_CHKN_BIT_REG, - I915_READ(GAMT_CHKN_BIT_REG) | - GAMT_CHKN_DISABLE_DYNAMIC_CREDIT_SHARING); + if (IS_KBL_REVID(i915, 0, KBL_REVID_B0)) + wa_write_or(wal, + GAMT_CHKN_BIT_REG, + GAMT_CHKN_DISABLE_DYNAMIC_CREDIT_SHARING); /* WaDisableGafsUnitClkGating:kbl */ - I915_WRITE(GEN7_UCGCTL4, - I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); + wa_write_or(wal, + GEN7_UCGCTL4, + GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); /* WaInPlaceDecompressionHang:kbl */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); - - /* WaKBLVECSSemaphoreWaitPoll:kbl */ - if (IS_KBL_REVID(dev_priv, KBL_REVID_A0, KBL_REVID_E0)) { - struct intel_engine_cs *engine; - unsigned int tmp; - - for_each_engine(engine, dev_priv, tmp) { - if (engine->id == RCS) - continue; - - I915_WRITE(RING_SEMA_WAIT_POLL(engine->mmio_base), 1); - } - } + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); } -static void glk_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void glk_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + gen9_gt_workarounds_init(i915); } -static void cfl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void cfl_gt_workarounds_init(struct drm_i915_private *i915) { - gen9_gt_workarounds_apply(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; - /* WaEnableGapsTsvCreditFix:cfl */ - I915_WRITE(GEN8_GARBCNTL, - I915_READ(GEN8_GARBCNTL) | GEN9_GAPS_TSV_CREDIT_DISABLE); + gen9_gt_workarounds_init(i915); /* WaDisableGafsUnitClkGating:cfl */ - I915_WRITE(GEN7_UCGCTL4, - I915_READ(GEN7_UCGCTL4) | GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); + wa_write_or(wal, + GEN7_UCGCTL4, + GEN8_EU_GAUNIT_CLOCK_GATE_DISABLE); /* WaInPlaceDecompressionHang:cfl */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); } static void wa_init_mcr(struct drm_i915_private *dev_priv) { const struct sseu_dev_info *sseu = &(INTEL_INFO(dev_priv)->sseu); - u32 mcr; + struct i915_wa_list *wal = &dev_priv->gt_wa_list; u32 mcr_slice_subslice_mask; /* @@ -770,8 +799,6 @@ static void wa_init_mcr(struct drm_i915_private *dev_priv) WARN_ON((enabled_mask & disabled_mask) != enabled_mask); } - mcr = I915_READ(GEN8_MCR_SELECTOR); - if (INTEL_GEN(dev_priv) >= 11) mcr_slice_subslice_mask = GEN11_MCR_SLICE_MASK | GEN11_MCR_SUBSLICE_MASK; @@ -789,148 +816,170 @@ static void wa_init_mcr(struct drm_i915_private *dev_priv) * occasions, such as INSTDONE, where this value is dependent * on s/ss combo, the read should be done with read_subslice_reg. */ - mcr &= ~mcr_slice_subslice_mask; - mcr |= intel_calculate_mcr_s_ss_select(dev_priv); - I915_WRITE(GEN8_MCR_SELECTOR, mcr); + wa_write_masked_or(wal, + GEN8_MCR_SELECTOR, + mcr_slice_subslice_mask, + intel_calculate_mcr_s_ss_select(dev_priv)); } -static void cnl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void cnl_gt_workarounds_init(struct drm_i915_private *i915) { - wa_init_mcr(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; + + wa_init_mcr(i915); /* WaDisableI2mCycleOnWRPort:cnl (pre-prod) */ - if (IS_CNL_REVID(dev_priv, CNL_REVID_B0, CNL_REVID_B0)) - I915_WRITE(GAMT_CHKN_BIT_REG, - I915_READ(GAMT_CHKN_BIT_REG) | - GAMT_CHKN_DISABLE_I2M_CYCLE_ON_WR_PORT); + if (IS_CNL_REVID(i915, CNL_REVID_B0, CNL_REVID_B0)) + wa_write_or(wal, + GAMT_CHKN_BIT_REG, + GAMT_CHKN_DISABLE_I2M_CYCLE_ON_WR_PORT); /* WaInPlaceDecompressionHang:cnl */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, - I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); - - /* WaEnablePreemptionGranularityControlByUMD:cnl */ - I915_WRITE(GEN7_FF_SLICE_CS_CHICKEN1, - _MASKED_BIT_ENABLE(GEN9_FFSC_PERCTX_PREEMPT_CTRL)); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); } -static void icl_gt_workarounds_apply(struct drm_i915_private *dev_priv) +static void icl_gt_workarounds_init(struct drm_i915_private *i915) { - wa_init_mcr(dev_priv); + struct i915_wa_list *wal = &i915->gt_wa_list; - /* This is not an Wa. Enable for better image quality */ - I915_WRITE(_3D_CHICKEN3, - _MASKED_BIT_ENABLE(_3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE)); + wa_init_mcr(i915); /* WaInPlaceDecompressionHang:icl */ - I915_WRITE(GEN9_GAMT_ECO_REG_RW_IA, I915_READ(GEN9_GAMT_ECO_REG_RW_IA) | - GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); - - /* WaPipelineFlushCoherentLines:icl */ - I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) | - GEN8_LQSC_FLUSH_COHERENT_LINES); - - /* Wa_1405543622:icl - * Formerly known as WaGAPZPriorityScheme - */ - I915_WRITE(GEN8_GARBCNTL, I915_READ(GEN8_GARBCNTL) | - GEN11_ARBITRATION_PRIO_ORDER_MASK); - - /* Wa_1604223664:icl - * Formerly known as WaL3BankAddressHashing - */ - I915_WRITE(GEN8_GARBCNTL, - (I915_READ(GEN8_GARBCNTL) & ~GEN11_HASH_CTRL_EXCL_MASK) | - GEN11_HASH_CTRL_EXCL_BIT0); - I915_WRITE(GEN11_GLBLINVL, - (I915_READ(GEN11_GLBLINVL) & ~GEN11_BANK_HASH_ADDR_EXCL_MASK) | - GEN11_BANK_HASH_ADDR_EXCL_BIT0); + wa_write_or(wal, + GEN9_GAMT_ECO_REG_RW_IA, + GAMT_ECO_ENABLE_IN_PLACE_DECOMPRESS); /* WaModifyGamTlbPartitioning:icl */ - I915_WRITE(GEN11_GACB_PERF_CTRL, - (I915_READ(GEN11_GACB_PERF_CTRL) & ~GEN11_HASH_CTRL_MASK) | - GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4); - - /* Wa_1405733216:icl - * Formerly known as WaDisableCleanEvicts - */ - I915_WRITE(GEN8_L3SQCREG4, I915_READ(GEN8_L3SQCREG4) | - GEN11_LQSC_CLEAN_EVICT_DISABLE); + wa_write_masked_or(wal, + GEN11_GACB_PERF_CTRL, + GEN11_HASH_CTRL_MASK, + GEN11_HASH_CTRL_BIT0 | GEN11_HASH_CTRL_BIT4); /* Wa_1405766107:icl * Formerly known as WaCL2SFHalfMaxAlloc */ - I915_WRITE(GEN11_LSN_UNSLCVC, I915_READ(GEN11_LSN_UNSLCVC) | - GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC | - GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC); + wa_write_or(wal, + GEN11_LSN_UNSLCVC, + GEN11_LSN_UNSLCVC_GAFS_HALF_SF_MAXALLOC | + GEN11_LSN_UNSLCVC_GAFS_HALF_CL2_MAXALLOC); /* Wa_220166154:icl * Formerly known as WaDisCtxReload */ - I915_WRITE(GAMW_ECO_DEV_RW_IA_REG, I915_READ(GAMW_ECO_DEV_RW_IA_REG) | - GAMW_ECO_DEV_CTX_RELOAD_DISABLE); + wa_write_or(wal, + GEN8_GAMW_ECO_DEV_RW_IA, + GAMW_ECO_DEV_CTX_RELOAD_DISABLE); /* Wa_1405779004:icl (pre-prod) */ - if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_A0)) - I915_WRITE(SLICE_UNIT_LEVEL_CLKGATE, - I915_READ(SLICE_UNIT_LEVEL_CLKGATE) | - MSCUNIT_CLKGATE_DIS); + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_A0)) + wa_write_or(wal, + SLICE_UNIT_LEVEL_CLKGATE, + MSCUNIT_CLKGATE_DIS); /* Wa_1406680159:icl */ - I915_WRITE(SUBSLICE_UNIT_LEVEL_CLKGATE, - I915_READ(SUBSLICE_UNIT_LEVEL_CLKGATE) | - GWUNIT_CLKGATE_DIS); - - /* Wa_1604302699:icl */ - I915_WRITE(GEN10_L3_CHICKEN_MODE_REGISTER, - I915_READ(GEN10_L3_CHICKEN_MODE_REGISTER) | - GEN11_I2M_WRITE_DISABLE); + wa_write_or(wal, + SUBSLICE_UNIT_LEVEL_CLKGATE, + GWUNIT_CLKGATE_DIS); /* Wa_1406838659:icl (pre-prod) */ - if (IS_ICL_REVID(dev_priv, ICL_REVID_A0, ICL_REVID_B0)) - I915_WRITE(INF_UNIT_LEVEL_CLKGATE, - I915_READ(INF_UNIT_LEVEL_CLKGATE) | - CGPSF_CLKGATE_DIS); - - /* WaForwardProgressSoftReset:icl */ - I915_WRITE(GEN10_SCRATCH_LNCF2, - I915_READ(GEN10_SCRATCH_LNCF2) | - PMFLUSHDONE_LNICRSDROP | - PMFLUSH_GAPL3UNBLOCK | - PMFLUSHDONE_LNEBLK); + if (IS_ICL_REVID(i915, ICL_REVID_A0, ICL_REVID_B0)) + wa_write_or(wal, + INF_UNIT_LEVEL_CLKGATE, + CGPSF_CLKGATE_DIS); /* Wa_1406463099:icl * Formerly known as WaGamTlbPendError */ - I915_WRITE(GAMT_CHKN_BIT_REG, - I915_READ(GAMT_CHKN_BIT_REG) | - GAMT_CHKN_DISABLE_L3_COH_PIPE); + wa_write_or(wal, + GAMT_CHKN_BIT_REG, + GAMT_CHKN_DISABLE_L3_COH_PIPE); } -void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv) +void intel_gt_init_workarounds(struct drm_i915_private *i915) { - if (INTEL_GEN(dev_priv) < 8) + struct i915_wa_list *wal = &i915->gt_wa_list; + + wa_init_start(wal, "GT"); + + if (INTEL_GEN(i915) < 8) return; - else if (IS_BROADWELL(dev_priv)) - bdw_gt_workarounds_apply(dev_priv); - else if (IS_CHERRYVIEW(dev_priv)) - chv_gt_workarounds_apply(dev_priv); - else if (IS_SKYLAKE(dev_priv)) - skl_gt_workarounds_apply(dev_priv); - else if (IS_BROXTON(dev_priv)) - bxt_gt_workarounds_apply(dev_priv); - else if (IS_KABYLAKE(dev_priv)) - kbl_gt_workarounds_apply(dev_priv); - else if (IS_GEMINILAKE(dev_priv)) - glk_gt_workarounds_apply(dev_priv); - else if (IS_COFFEELAKE(dev_priv)) - cfl_gt_workarounds_apply(dev_priv); - else if (IS_CANNONLAKE(dev_priv)) - cnl_gt_workarounds_apply(dev_priv); - else if (IS_ICELAKE(dev_priv)) - icl_gt_workarounds_apply(dev_priv); + else if (IS_BROADWELL(i915)) + return; + else if (IS_CHERRYVIEW(i915)) + return; + else if (IS_SKYLAKE(i915)) + skl_gt_workarounds_init(i915); + else if (IS_BROXTON(i915)) + bxt_gt_workarounds_init(i915); + else if (IS_KABYLAKE(i915)) + kbl_gt_workarounds_init(i915); + else if (IS_GEMINILAKE(i915)) + glk_gt_workarounds_init(i915); + else if (IS_COFFEELAKE(i915)) + cfl_gt_workarounds_init(i915); + else if (IS_CANNONLAKE(i915)) + cnl_gt_workarounds_init(i915); + else if (IS_ICELAKE(i915)) + icl_gt_workarounds_init(i915); else - MISSING_CASE(INTEL_GEN(dev_priv)); + MISSING_CASE(INTEL_GEN(i915)); + + wa_init_finish(wal); +} + +static enum forcewake_domains +wal_get_fw_for_rmw(struct drm_i915_private *dev_priv, + const struct i915_wa_list *wal) +{ + enum forcewake_domains fw = 0; + struct i915_wa *wa; + unsigned int i; + + for (i = 0, wa = wal->list; i < wal->count; i++, wa++) + fw |= intel_uncore_forcewake_for_reg(dev_priv, + wa->reg, + FW_REG_READ | + FW_REG_WRITE); + + return fw; +} + +static void +wa_list_apply(struct drm_i915_private *dev_priv, const struct i915_wa_list *wal) +{ + enum forcewake_domains fw; + unsigned long flags; + struct i915_wa *wa; + unsigned int i; + + if (!wal->count) + return; + + fw = wal_get_fw_for_rmw(dev_priv, wal); + + spin_lock_irqsave(&dev_priv->uncore.lock, flags); + intel_uncore_forcewake_get__locked(dev_priv, fw); + + for (i = 0, wa = wal->list; i < wal->count; i++, wa++) { + u32 val = I915_READ_FW(wa->reg); + + val &= ~wa->mask; + val |= wa->val; + + I915_WRITE_FW(wa->reg, val); + } + + intel_uncore_forcewake_put__locked(dev_priv, fw); + spin_unlock_irqrestore(&dev_priv->uncore.lock, flags); + + DRM_DEBUG_DRIVER("Applied %u %s workarounds\n", wal->count, wal->name); +} + +void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv) +{ + wa_list_apply(dev_priv, &dev_priv->gt_wa_list); } struct whitelist { @@ -1077,6 +1126,146 @@ void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine) whitelist_apply(engine, whitelist_build(engine, &w)); } +static void rcs_engine_wa_init(struct intel_engine_cs *engine) +{ + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->wa_list; + + if (IS_ICELAKE(i915)) { + /* This is not an Wa. Enable for better image quality */ + wa_masked_en(wal, + _3D_CHICKEN3, + _3D_CHICKEN3_AA_LINE_QUALITY_FIX_ENABLE); + + /* WaPipelineFlushCoherentLines:icl */ + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN8_LQSC_FLUSH_COHERENT_LINES); + + /* + * Wa_1405543622:icl + * Formerly known as WaGAPZPriorityScheme + */ + wa_write_or(wal, + GEN8_GARBCNTL, + GEN11_ARBITRATION_PRIO_ORDER_MASK); + + /* + * Wa_1604223664:icl + * Formerly known as WaL3BankAddressHashing + */ + wa_write_masked_or(wal, + GEN8_GARBCNTL, + GEN11_HASH_CTRL_EXCL_MASK, + GEN11_HASH_CTRL_EXCL_BIT0); + wa_write_masked_or(wal, + GEN11_GLBLINVL, + GEN11_BANK_HASH_ADDR_EXCL_MASK, + GEN11_BANK_HASH_ADDR_EXCL_BIT0); + + /* + * Wa_1405733216:icl + * Formerly known as WaDisableCleanEvicts + */ + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN11_LQSC_CLEAN_EVICT_DISABLE); + + /* Wa_1604302699:icl */ + wa_write_or(wal, + GEN10_L3_CHICKEN_MODE_REGISTER, + GEN11_I2M_WRITE_DISABLE); + + /* WaForwardProgressSoftReset:icl */ + wa_write_or(wal, + GEN10_SCRATCH_LNCF2, + PMFLUSHDONE_LNICRSDROP | + PMFLUSH_GAPL3UNBLOCK | + PMFLUSHDONE_LNEBLK); + } + + if (IS_GEN9(i915) || IS_CANNONLAKE(i915)) { + /* WaEnablePreemptionGranularityControlByUMD:skl,bxt,kbl,cfl,cnl */ + wa_masked_en(wal, + GEN7_FF_SLICE_CS_CHICKEN1, + GEN9_FFSC_PERCTX_PREEMPT_CTRL); + } + + if (IS_SKYLAKE(i915) || IS_KABYLAKE(i915) || IS_COFFEELAKE(i915)) { + /* WaEnableGapsTsvCreditFix:skl,kbl,cfl */ + wa_write_or(wal, + GEN8_GARBCNTL, + GEN9_GAPS_TSV_CREDIT_DISABLE); + } + + if (IS_BROXTON(i915)) { + /* WaDisablePooledEuLoadBalancingFix:bxt */ + wa_masked_en(wal, + FF_SLICE_CS_CHICKEN2, + GEN9_POOLED_EU_LOAD_BALANCING_FIX_DISABLE); + } + + if (IS_GEN9(i915)) { + /* WaContextSwitchWithConcurrentTLBInvalidate:skl,bxt,kbl,glk,cfl */ + wa_masked_en(wal, + GEN9_CSFE_CHICKEN1_RCS, + GEN9_PREEMPT_GPGPU_SYNC_SWITCH_DISABLE); + + /* WaEnableLbsSlaRetryTimerDecrement:skl,bxt,kbl,glk,cfl */ + wa_write_or(wal, + BDW_SCRATCH1, + GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE); + + /* WaProgramL3SqcReg1DefaultForPerf:bxt,glk */ + if (IS_GEN9_LP(i915)) + wa_write_masked_or(wal, + GEN8_L3SQCREG1, + L3_PRIO_CREDITS_MASK, + L3_GENERAL_PRIO_CREDITS(62) | + L3_HIGH_PRIO_CREDITS(2)); + + /* WaOCLCoherentLineFlush:skl,bxt,kbl,cfl */ + wa_write_or(wal, + GEN8_L3SQCREG4, + GEN8_LQSC_FLUSH_COHERENT_LINES); + } +} + +static void xcs_engine_wa_init(struct intel_engine_cs *engine) +{ + struct drm_i915_private *i915 = engine->i915; + struct i915_wa_list *wal = &engine->wa_list; + + /* WaKBLVECSSemaphoreWaitPoll:kbl */ + if (IS_KBL_REVID(i915, KBL_REVID_A0, KBL_REVID_E0)) { + wa_write(wal, + RING_SEMA_WAIT_POLL(engine->mmio_base), + 1); + } +} + +void intel_engine_init_workarounds(struct intel_engine_cs *engine) +{ + struct i915_wa_list *wal = &engine->wa_list; + + if (GEM_WARN_ON(INTEL_GEN(engine->i915) < 8)) + return; + + wa_init_start(wal, engine->name); + + if (engine->id == RCS) + rcs_engine_wa_init(engine); + else + xcs_engine_wa_init(engine); + + wa_init_finish(wal); +} + +void intel_engine_apply_workarounds(struct intel_engine_cs *engine) +{ + wa_list_apply(engine->i915, &engine->wa_list); +} + #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST) #include "selftests/intel_workarounds.c" #endif diff --git a/drivers/gpu/drm/i915/intel_workarounds.h b/drivers/gpu/drm/i915/intel_workarounds.h index b11d0623e626..979695a53964 100644 --- a/drivers/gpu/drm/i915/intel_workarounds.h +++ b/drivers/gpu/drm/i915/intel_workarounds.h @@ -7,11 +7,35 @@ #ifndef _I915_WORKAROUNDS_H_ #define _I915_WORKAROUNDS_H_ +#include <linux/slab.h> + +struct i915_wa { + i915_reg_t reg; + u32 mask; + u32 val; +}; + +struct i915_wa_list { + const char *name; + struct i915_wa *list; + unsigned int count; +}; + +static inline void intel_wa_list_free(struct i915_wa_list *wal) +{ + kfree(wal->list); + memset(wal, 0, sizeof(*wal)); +} + int intel_ctx_workarounds_init(struct drm_i915_private *dev_priv); int intel_ctx_workarounds_emit(struct i915_request *rq); -void intel_gt_workarounds_apply(struct drm_i915_private *dev_priv); +void intel_gt_init_workarounds(struct drm_i915_private *dev_priv); +void intel_gt_apply_workarounds(struct drm_i915_private *dev_priv); void intel_whitelist_workarounds_apply(struct intel_engine_cs *engine); +void intel_engine_init_workarounds(struct intel_engine_cs *engine); +void intel_engine_apply_workarounds(struct intel_engine_cs *engine); + #endif diff --git a/drivers/gpu/drm/i915/selftests/huge_pages.c b/drivers/gpu/drm/i915/selftests/huge_pages.c index 8d03f64eabd7..5c22f2c8d4cf 100644 --- a/drivers/gpu/drm/i915/selftests/huge_pages.c +++ b/drivers/gpu/drm/i915/selftests/huge_pages.c @@ -551,7 +551,7 @@ static int igt_mock_ppgtt_misaligned_dma(void *arg) err = igt_check_page_sizes(vma); if (vma->page_sizes.gtt != I915_GTT_PAGE_SIZE_4K) { - pr_err("page_sizes.gtt=%u, expected %lu\n", + pr_err("page_sizes.gtt=%u, expected %llu\n", vma->page_sizes.gtt, I915_GTT_PAGE_SIZE_4K); err = -EINVAL; } diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c index 8e2e269db97e..127d81513671 100644 --- a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c @@ -1337,7 +1337,7 @@ static int igt_gtt_reserve(void *arg) GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); if (vma->node.start != total || vma->node.size != 2*I915_GTT_PAGE_SIZE) { - pr_err("i915_gem_gtt_reserve (pass 1) placement failed, found (%llx + %llx), expected (%llx + %lx)\n", + pr_err("i915_gem_gtt_reserve (pass 1) placement failed, found (%llx + %llx), expected (%llx + %llx)\n", vma->node.start, vma->node.size, total, 2*I915_GTT_PAGE_SIZE); err = -EINVAL; @@ -1386,7 +1386,7 @@ static int igt_gtt_reserve(void *arg) GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); if (vma->node.start != total || vma->node.size != 2*I915_GTT_PAGE_SIZE) { - pr_err("i915_gem_gtt_reserve (pass 2) placement failed, found (%llx + %llx), expected (%llx + %lx)\n", + pr_err("i915_gem_gtt_reserve (pass 2) placement failed, found (%llx + %llx), expected (%llx + %llx)\n", vma->node.start, vma->node.size, total, 2*I915_GTT_PAGE_SIZE); err = -EINVAL; @@ -1430,7 +1430,7 @@ static int igt_gtt_reserve(void *arg) GEM_BUG_ON(!drm_mm_node_allocated(&vma->node)); if (vma->node.start != offset || vma->node.size != 2*I915_GTT_PAGE_SIZE) { - pr_err("i915_gem_gtt_reserve (pass 3) placement failed, found (%llx + %llx), expected (%llx + %lx)\n", + pr_err("i915_gem_gtt_reserve (pass 3) placement failed, found (%llx + %llx), expected (%llx + %llx)\n", vma->node.start, vma->node.size, offset, 2*I915_GTT_PAGE_SIZE); err = -EINVAL; |