diff options
| author | Daniel Vetter <daniel.vetter@ffwll.ch> | 2017-07-26 13:23:10 +0200 |
|---|---|---|
| committer | Daniel Vetter <daniel.vetter@ffwll.ch> | 2017-07-26 13:43:33 +0200 |
| commit | af055598542670c8533a58582813b1419949cae0 (patch) | |
| tree | 901fa1bf635d5c1e91d08f9f4c2e4943516dbb71 /drivers/gpu/drm | |
| parent | 9f15a4ab19ab33658dbc9fd37be5210e8c1ac622 (diff) | |
| parent | 2d62c799f8ffac4f7ffba6a4e7f148827dfc24c7 (diff) | |
| download | linux-af055598542670c8533a58582813b1419949cae0.tar.gz linux-af055598542670c8533a58582813b1419949cae0.tar.bz2 linux-af055598542670c8533a58582813b1419949cae0.zip | |
Merge airlied/drm-next into drm-misc-next
I need this to be able to apply the deferred fbdev setup patches, I
need the relevant prep work that landed through the drm-intel tree.
Also squash in conflict fixup from Laurent Pinchart.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Diffstat (limited to 'drivers/gpu/drm')
137 files changed, 3365 insertions, 2318 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 12d61edb3597..ff7bf1a9f967 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -1028,12 +1028,15 @@ struct amdgpu_gfx_config { }; struct amdgpu_cu_info { - uint32_t number; /* total active CU number */ - uint32_t ao_cu_mask; uint32_t max_waves_per_simd; uint32_t wave_front_size; uint32_t max_scratch_slots_per_cu; uint32_t lds_size; + + /* total active CU number */ + uint32_t number; + uint32_t ao_cu_mask; + uint32_t ao_cu_bitmap[4][4]; uint32_t bitmap[4][4]; }; @@ -1924,7 +1927,6 @@ void amdgpu_pci_config_reset(struct amdgpu_device *adev); bool amdgpu_need_post(struct amdgpu_device *adev); void amdgpu_update_display_priority(struct amdgpu_device *adev); -int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data); void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev, u64 num_bytes); void amdgpu_ttm_placement_from_domain(struct amdgpu_bo *abo, u32 domain); bool amdgpu_ttm_bo_is_amdgpu_bo(struct ttm_buffer_object *bo); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c index c6dba1eaefbd..c0a806280257 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cgs.c @@ -838,6 +838,12 @@ static int amdgpu_cgs_get_active_displays_info(struct cgs_device *cgs_device, return -EINVAL; mode_info = info->mode_info; + if (mode_info) { + /* if the displays are off, vblank time is max */ + mode_info->vblank_time_us = 0xffffffff; + /* always set the reference clock */ + mode_info->ref_clock = adev->clock.spll.reference_freq; + } if (adev->mode_info.num_crtc && adev->mode_info.mode_config_initialized) { list_for_each_entry(crtc, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index aeee6840e82b..5599c01b265d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -64,7 +64,7 @@ static int amdgpu_cs_user_fence_chunk(struct amdgpu_cs_parser *p, return 0; } -int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) +static int amdgpu_cs_parser_init(struct amdgpu_cs_parser *p, void *data) { struct amdgpu_fpriv *fpriv = p->filp->driver_priv; struct amdgpu_vm *vm = &fpriv->vm; @@ -497,7 +497,7 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p, &e->user_invalidated) && e->user_pages) { /* We acquired a page array, but somebody - * invalidated it. Free it an try again + * invalidated it. Free it and try again */ release_pages(e->user_pages, e->robj->tbo.ttm->num_pages, @@ -1069,10 +1069,8 @@ static void amdgpu_cs_post_dependencies(struct amdgpu_cs_parser *p) { int i; - for (i = 0; i < p->num_post_dep_syncobjs; ++i) { - drm_syncobj_replace_fence(p->filp, p->post_dep_syncobjs[i], - p->fence); - } + for (i = 0; i < p->num_post_dep_syncobjs; ++i) + drm_syncobj_replace_fence(p->post_dep_syncobjs[i], p->fence); } static int amdgpu_cs_submit(struct amdgpu_cs_parser *p, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index b2c960b2ea82..4a8fc15467cf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -1162,16 +1162,12 @@ static void amdgpu_switcheroo_set_state(struct pci_dev *pdev, enum vga_switchero return; if (state == VGA_SWITCHEROO_ON) { - unsigned d3_delay = dev->pdev->d3_delay; - pr_info("amdgpu: switched on\n"); /* don't suspend or resume card normally */ dev->switch_power_state = DRM_SWITCH_POWER_CHANGING; amdgpu_device_resume(dev, true, true); - dev->pdev->d3_delay = d3_delay; - dev->switch_power_state = DRM_SWITCH_POWER_ON; drm_kms_helper_poll_enable(dev); } else { @@ -3804,7 +3800,7 @@ int amdgpu_debugfs_init(struct drm_minor *minor) return 0; } #else -static int amdgpu_debugfs_test_ib_init(struct amdgpu_device *adev) +static int amdgpu_debugfs_test_ib_ring_init(struct amdgpu_device *adev) { return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 0148dd32e561..469992470953 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -67,9 +67,10 @@ * - 3.15.0 - Export more gpu info for gfx9 * - 3.16.0 - Add reserved vmid support * - 3.17.0 - Add AMDGPU_NUM_VRAM_CPU_PAGE_FAULTS. + * - 3.18.0 - Export gpu always on cu bitmap */ #define KMS_DRIVER_MAJOR 3 -#define KMS_DRIVER_MINOR 17 +#define KMS_DRIVER_MINOR 18 #define KMS_DRIVER_PATCHLEVEL 0 int amdgpu_vram_limit = 0; @@ -247,14 +248,28 @@ MODULE_PARM_DESC(lbpw, "Load Balancing Per Watt (LBPW) support (1 = enable, 0 = module_param_named(lbpw, amdgpu_lbpw, int, 0444); #ifdef CONFIG_DRM_AMDGPU_SI + +#if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE) int amdgpu_si_support = 0; MODULE_PARM_DESC(si_support, "SI support (1 = enabled, 0 = disabled (default))"); +#else +int amdgpu_si_support = 1; +MODULE_PARM_DESC(si_support, "SI support (1 = enabled (default), 0 = disabled)"); +#endif + module_param_named(si_support, amdgpu_si_support, int, 0444); #endif #ifdef CONFIG_DRM_AMDGPU_CIK + +#if defined(CONFIG_DRM_RADEON) || defined(CONFIG_DRM_RADEON_MODULE) int amdgpu_cik_support = 0; MODULE_PARM_DESC(cik_support, "CIK support (1 = enabled, 0 = disabled (default))"); +#else +int amdgpu_cik_support = 1; +MODULE_PARM_DESC(cik_support, "CIK support (1 = enabled (default), 0 = disabled)"); +#endif + module_param_named(cik_support, amdgpu_cik_support, int, 0444); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index 12497a40ef92..b0b23101d1c8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -594,6 +594,8 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file dev_info.cu_active_number = adev->gfx.cu_info.number; dev_info.cu_ao_mask = adev->gfx.cu_info.ao_cu_mask; dev_info.ce_ram_size = adev->gfx.ce_ram_size; + memcpy(&dev_info.cu_ao_bitmap[0], &adev->gfx.cu_info.ao_cu_bitmap[0], + sizeof(adev->gfx.cu_info.ao_cu_bitmap)); memcpy(&dev_info.cu_bitmap[0], &adev->gfx.cu_info.bitmap[0], sizeof(adev->gfx.cu_info.bitmap)); dev_info.vram_type = adev->mc.vram_type; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c index 72c03c744594..b7e1c026c0c8 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_powerplay.c @@ -188,6 +188,9 @@ static int amdgpu_pp_hw_fini(void *handle) int ret = 0; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (adev->pp_enabled && adev->pm.dpm_enabled) + amdgpu_pm_sysfs_fini(adev); + if (adev->powerplay.ip_funcs->hw_fini) ret = adev->powerplay.ip_funcs->hw_fini( adev->powerplay.pp_handle); @@ -206,10 +209,9 @@ static void amdgpu_pp_late_fini(void *handle) adev->powerplay.ip_funcs->late_fini( adev->powerplay.pp_handle); - if (adev->pp_enabled && adev->pm.dpm_enabled) - amdgpu_pm_sysfs_fini(adev); - amd_powerplay_destroy(adev->powerplay.pp_handle); + if (adev->pp_enabled) + amd_powerplay_destroy(adev->powerplay.pp_handle); } static int amdgpu_pp_suspend(void *handle) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index c224c5caba5b..4083be61b328 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -152,8 +152,8 @@ static void psp_prep_tmr_cmd_buf(struct psp_gfx_cmd_resp *cmd, uint64_t tmr_mc, uint32_t size) { cmd->cmd_id = GFX_CMD_ID_SETUP_TMR; - cmd->cmd.cmd_setup_tmr.buf_phy_addr_lo = (uint32_t)tmr_mc; - cmd->cmd.cmd_setup_tmr.buf_phy_addr_hi = (uint32_t)(tmr_mc >> 32); + cmd->cmd.cmd_setup_tmr.buf_phy_addr_lo = lower_32_bits(tmr_mc); + cmd->cmd.cmd_setup_tmr.buf_phy_addr_hi = upper_32_bits(tmr_mc); cmd->cmd.cmd_setup_tmr.buf_size = size; } @@ -333,14 +333,11 @@ static int psp_load_fw(struct amdgpu_device *adev) { int ret; struct psp_context *psp = &adev->psp; - struct psp_gfx_cmd_resp *cmd; - cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); - if (!cmd) + psp->cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL); + if (!psp->cmd) return -ENOMEM; - psp->cmd = cmd; - ret = amdgpu_bo_create_kernel(adev, PSP_1_MEG, PSP_1_MEG, AMDGPU_GEM_DOMAIN_GTT, &psp->fw_pri_bo, @@ -379,8 +376,6 @@ static int psp_load_fw(struct amdgpu_device *adev) if (ret) goto failed_mem; - kfree(cmd); - return 0; failed_mem: @@ -390,7 +385,8 @@ failed_mem1: amdgpu_bo_free_kernel(&psp->fw_pri_bo, &psp->fw_pri_mc_addr, &psp->fw_pri_buf); failed: - kfree(cmd); + kfree(psp->cmd); + psp->cmd = NULL; return ret; } @@ -450,6 +446,9 @@ static int psp_hw_fini(void *handle) amdgpu_bo_free_kernel(&psp->fence_buf_bo, &psp->fence_buf_mc_addr, &psp->fence_buf); + kfree(psp->cmd); + psp->cmd = NULL; + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c index 7b0b3cf16334..5173ca1fd159 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c @@ -3535,7 +3535,9 @@ static void gfx_v6_0_get_cu_info(struct amdgpu_device *adev) mask <<= 1; } active_cu_number += counter; - ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + if (i < 2 && j < 2) + ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + cu_info->ao_cu_bitmap[i][j] = ao_bitmap; } } diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c index ec754288f146..37b45e4403d1 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v7_0.c @@ -5427,7 +5427,9 @@ static void gfx_v7_0_get_cu_info(struct amdgpu_device *adev) mask <<= 1; } active_cu_number += counter; - ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + if (i < 2 && j < 2) + ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + cu_info->ao_cu_bitmap[i][j] = ao_bitmap; } } gfx_v7_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c index 142924212b43..aa5a50f5eac8 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c @@ -40,7 +40,6 @@ #include "bif/bif_5_0_d.h" #include "bif/bif_5_0_sh_mask.h" - #include "gca/gfx_8_0_d.h" #include "gca/gfx_8_0_enum.h" #include "gca/gfx_8_0_sh_mask.h" @@ -2100,7 +2099,7 @@ static int gfx_v8_0_sw_init(void *handle) return r; /* create MQD for all compute queues as well as KIQ for SRIOV case */ - r = amdgpu_gfx_compute_mqd_sw_init(adev, sizeof(struct vi_mqd)); + r = amdgpu_gfx_compute_mqd_sw_init(adev, sizeof(struct vi_mqd_allocation)); if (r) return r; @@ -4637,56 +4636,6 @@ static int gfx_v8_0_kiq_kcq_enable(struct amdgpu_device *adev) return r; } -static int gfx_v8_0_kiq_kcq_disable(struct amdgpu_device *adev) -{ - struct amdgpu_ring *kiq_ring = &adev->gfx.kiq.ring; - uint32_t scratch, tmp = 0; - int r, i; - - r = amdgpu_gfx_scratch_get(adev, &scratch); - if (r) { - DRM_ERROR("Failed to get scratch reg (%d).\n", r); - return r; - } - WREG32(scratch, 0xCAFEDEAD); - - r = amdgpu_ring_alloc(kiq_ring, 6 + 3); - if (r) { - DRM_ERROR("Failed to lock KIQ (%d).\n", r); - amdgpu_gfx_scratch_free(adev, scratch); - return r; - } - /* unmap queues */ - amdgpu_ring_write(kiq_ring, PACKET3(PACKET3_UNMAP_QUEUES, 4)); - amdgpu_ring_write(kiq_ring, - PACKET3_UNMAP_QUEUES_ACTION(1)| /* RESET_QUEUES */ - PACKET3_UNMAP_QUEUES_QUEUE_SEL(2)); /* select all queues */ - amdgpu_ring_write(kiq_ring, 0); - amdgpu_ring_write(kiq_ring, 0); - amdgpu_ring_write(kiq_ring, 0); - amdgpu_ring_write(kiq_ring, 0); - /* write to scratch for completion */ - amdgpu_ring_write(kiq_ring, PACKET3(PACKET3_SET_UCONFIG_REG, 1)); - amdgpu_ring_write(kiq_ring, (scratch - PACKET3_SET_UCONFIG_REG_START)); - amdgpu_ring_write(kiq_ring, 0xDEADBEEF); - amdgpu_ring_commit(kiq_ring); - - for (i = 0; i < adev->usec_timeout; i++) { - tmp = RREG32(scratch); - if (tmp == 0xDEADBEEF) - break; - DRM_UDELAY(1); - } - if (i >= adev->usec_timeout) { - DRM_ERROR("KCQ disabled failed (scratch(0x%04X)=0x%08X)\n", - scratch, tmp); - r = -EINVAL; - } - amdgpu_gfx_scratch_free(adev, scratch); - - return r; -} - static int gfx_v8_0_deactivate_hqd(struct amdgpu_device *adev, u32 req) { int i, r = 0; @@ -4715,9 +4664,6 @@ static int gfx_v8_0_mqd_init(struct amdgpu_ring *ring) uint64_t hqd_gpu_addr, wb_gpu_addr, eop_base_addr; uint32_t tmp; - /* init the mqd struct */ - memset(mqd, 0, sizeof(struct vi_mqd)); - mqd->header = 0xC0310800; mqd->compute_pipelinestat_enable = 0x00000001; mqd->compute_static_thread_mgmt_se0 = 0xffffffff; @@ -4725,7 +4671,12 @@ static int gfx_v8_0_mqd_init(struct amdgpu_ring *ring) mqd->compute_static_thread_mgmt_se2 = 0xffffffff; mqd->compute_static_thread_mgmt_se3 = 0xffffffff; mqd->compute_misc_reserved = 0x00000003; - + if (!(adev->flags & AMD_IS_APU)) { + mqd->dynamic_cu_mask_addr_lo = lower_32_bits(ring->mqd_gpu_addr + + offsetof(struct vi_mqd_allocation, dyamic_cu_mask)); + mqd->dynamic_cu_mask_addr_hi = upper_32_bits(ring->mqd_gpu_addr + + offsetof(struct vi_mqd_allocation, dyamic_cu_mask)); + } eop_base_addr = ring->eop_gpu_addr >> 8; mqd->cp_hqd_eop_base_addr_lo = eop_base_addr; mqd->cp_hqd_eop_base_addr_hi = upper_32_bits(eop_base_addr); @@ -4890,7 +4841,6 @@ int gfx_v8_0_mqd_commit(struct amdgpu_device *adev, static int gfx_v8_0_kiq_init_queue(struct amdgpu_ring *ring) { - int r = 0; struct amdgpu_device *adev = ring->adev; struct vi_mqd *mqd = ring->mqd_ptr; int mqd_idx = AMDGPU_MAX_COMPUTE_RINGS; @@ -4900,44 +4850,32 @@ static int gfx_v8_0_kiq_init_queue(struct amdgpu_ring *ring) if (adev->gfx.in_reset) { /* for GPU_RESET case */ /* reset MQD to a clean status */ if (adev->gfx.mec.mqd_backup[mqd_idx]) - memcpy(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(*mqd)); + memcpy(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(struct vi_mqd_allocation)); /* reset ring buffer */ ring->wptr = 0; amdgpu_ring_clear_ring(ring); mutex_lock(&adev->srbm_mutex); vi_srbm_select(adev, ring->me, ring->pipe, ring->queue, 0); - r = gfx_v8_0_deactivate_hqd(adev, 1); - if (r) { - dev_err(adev->dev, "failed to deactivate ring %s\n", ring->name); - goto out_unlock; - } gfx_v8_0_mqd_commit(adev, mqd); vi_srbm_select(adev, 0, 0, 0, 0); mutex_unlock(&adev->srbm_mutex); } else { + memset((void *)mqd, 0, sizeof(struct vi_mqd_allocation)); + ((struct vi_mqd_allocation *)mqd)->dyamic_cu_mask = 0xFFFFFFFF; + ((struct vi_mqd_allocation *)mqd)->dyamic_rb_mask = 0xFFFFFFFF; mutex_lock(&adev->srbm_mutex); vi_srbm_select(adev, ring->me, ring->pipe, ring->queue, 0); gfx_v8_0_mqd_init(ring); - r = gfx_v8_0_deactivate_hqd(adev, 1); - if (r) { - dev_err(adev->dev, "failed to deactivate ring %s\n", ring->name); - goto out_unlock; - } gfx_v8_0_mqd_commit(adev, mqd); vi_srbm_select(adev, 0, 0, 0, 0); mutex_unlock(&adev->srbm_mutex); if (adev->gfx.mec.mqd_backup[mqd_idx]) - memcpy(adev->gfx.mec.mqd_backup[mqd_idx], mqd, sizeof(*mqd)); + memcpy(adev->gfx.mec.mqd_backup[mqd_idx], mqd, sizeof(struct vi_mqd_allocation)); } - return r; - -out_unlock: - vi_srbm_select(adev, 0, 0, 0, 0); - mutex_unlock(&adev->srbm_mutex); - return r; + return 0; } static int gfx_v8_0_kcq_init_queue(struct amdgpu_ring *ring) @@ -4947,6 +4885,9 @@ static int gfx_v8_0_kcq_init_queue(struct amdgpu_ring *ring) int mqd_idx = ring - &adev->gfx.compute_ring[0]; if (!adev->gfx.in_reset && !adev->gfx.in_suspend) { + memset((void *)mqd, 0, sizeof(struct vi_mqd_allocation)); + ((struct vi_mqd_allocation *)mqd)->dyamic_cu_mask = 0xFFFFFFFF; + ((struct vi_mqd_allocation *)mqd)->dyamic_rb_mask = 0xFFFFFFFF; mutex_lock(&adev->srbm_mutex); vi_srbm_select(adev, ring->me, ring->pipe, ring->queue, 0); gfx_v8_0_mqd_init(ring); @@ -4954,11 +4895,11 @@ static int gfx_v8_0_kcq_init_queue(struct amdgpu_ring *ring) mutex_unlock(&adev->srbm_mutex); if (adev->gfx.mec.mqd_backup[mqd_idx]) - memcpy(adev->gfx.mec.mqd_backup[mqd_idx], mqd, sizeof(*mqd)); + memcpy(adev->gfx.mec.mqd_backup[mqd_idx], mqd, sizeof(struct vi_mqd_allocation)); } else if (adev->gfx.in_reset) { /* for GPU_RESET case */ /* reset MQD to a clean status */ if (adev->gfx.mec.mqd_backup[mqd_idx]) - memcpy(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(*mqd)); + memcpy(mqd, adev->gfx.mec.mqd_backup[mqd_idx], sizeof(struct vi_mqd_allocation)); /* reset ring buffer */ ring->wptr = 0; amdgpu_ring_clear_ring(ring); @@ -5138,7 +5079,6 @@ static int gfx_v8_0_hw_fini(void *handle) pr_debug("For SRIOV client, shouldn't do anything.\n"); return 0; } - gfx_v8_0_kiq_kcq_disable(adev); gfx_v8_0_cp_enable(adev, false); gfx_v8_0_rlc_stop(adev); @@ -7080,7 +7020,9 @@ static void gfx_v8_0_get_cu_info(struct amdgpu_device *adev) mask <<= 1; } active_cu_number += counter; - ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + if (i < 2 && j < 2) + ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + cu_info->ao_cu_bitmap[i][j] = ao_bitmap; } } gfx_v8_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index ba228f613027..3a0b69b09ed6 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -1964,8 +1964,8 @@ static void gfx_v9_0_enable_gfx_pipeline_powergating(struct amdgpu_device *adev, data = RREG32(SOC15_REG_OFFSET(GC, 0, mmDB_RENDER_CONTROL)); } -void gfx_v9_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev, - bool enable) +static void gfx_v9_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev, + bool enable) { uint32_t data, default_data; @@ -1978,7 +1978,7 @@ void gfx_v9_0_enable_gfx_static_mg_power_gating(struct amdgpu_device *adev, WREG32(SOC15_REG_OFFSET(GC, 0, mmRLC_PG_CNTL), data); } -void gfx_v9_0_enable_gfx_dynamic_mg_power_gating(struct amdgpu_device *adev, +static void gfx_v9_0_enable_gfx_dynamic_mg_power_gating(struct amdgpu_device *adev, bool enable) { uint32_t data, default_data; @@ -2502,56 +2502,6 @@ static int gfx_v9_0_kiq_kcq_enable(struct amdgpu_device *adev) return r; } -static int gfx_v9_0_kiq_kcq_disable(struct amdgpu_device *adev) -{ - struct amdgpu_ring *kiq_ring = &adev->gfx.kiq.ring; - uint32_t scratch, tmp = 0; - int r, i; - - r = amdgpu_gfx_scratch_get(adev, &scratch); - if (r) { - DRM_ERROR("Failed to get scratch reg (%d).\n", r); - return r; - } - WREG32(scratch, 0xCAFEDEAD); - - r = amdgpu_ring_alloc(kiq_ring, 6 + 3); - if (r) { - DRM_ERROR("Failed to lock KIQ (%d).\n", r); - amdgpu_gfx_scratch_free(adev, scratch); - return r; - } - /* unmap queues */ - amdgpu_ring_write(kiq_ring, PACKET3(PACKET3_UNMAP_QUEUES, 4)); - amdgpu_ring_write(kiq_ring, - PACKET3_UNMAP_QUEUES_ACTION(1)| /* RESET_QUEUES */ - PACKET3_UNMAP_QUEUES_QUEUE_SEL(2)); /* select all queues */ - amdgpu_ring_write(kiq_ring, 0); - amdgpu_ring_write(kiq_ring, 0); - amdgpu_ring_write(kiq_ring, 0); - amdgpu_ring_write(kiq_ring, 0); - /* write to scratch for completion */ - amdgpu_ring_write(kiq_ring, PACKET3(PACKET3_SET_UCONFIG_REG, 1)); - amdgpu_ring_write(kiq_ring, (scratch - PACKET3_SET_UCONFIG_REG_START)); - amdgpu_ring_write(kiq_ring, 0xDEADBEEF); - amdgpu_ring_commit(kiq_ring); - - for (i = 0; i < adev->usec_timeout; i++) { - tmp = RREG32(scratch); - if (tmp == 0xDEADBEEF) - break; - DRM_UDELAY(1); - } - if (i >= adev->usec_timeout) { - DRM_ERROR("KCQ disable failed (scratch(0x%04X)=0x%08X)\n", - scratch, tmp); - r = -EINVAL; - } - amdgpu_gfx_scratch_free(adev, scratch); - - return r; -} - static int gfx_v9_0_mqd_init(struct amdgpu_ring *ring) { struct amdgpu_device *adev = ring->adev; @@ -2996,7 +2946,6 @@ static int gfx_v9_0_hw_fini(void *handle) pr_debug("For SRIOV client, shouldn't do anything.\n"); return 0; } - gfx_v9_0_kiq_kcq_disable(adev); gfx_v9_0_cp_enable(adev, false); gfx_v9_0_rlc_stop(adev); @@ -4416,6 +4365,20 @@ static void gfx_v9_0_set_gds_init(struct amdgpu_device *adev) } } +static void gfx_v9_0_set_user_cu_inactive_bitmap(struct amdgpu_device *adev, + u32 bitmap) +{ + u32 data; + + if (!bitmap) + return; + + data = bitmap << GC_USER_SHADER_ARRAY_CONFIG__INACTIVE_CUS__SHIFT; + data &= GC_USER_SHADER_ARRAY_CONFIG__INACTIVE_CUS_MASK; + + WREG32_SOC15(GC, 0, mmGC_USER_SHADER_ARRAY_CONFIG, data); +} + static u32 gfx_v9_0_get_cu_active_bitmap(struct amdgpu_device *adev) { u32 data, mask; @@ -4436,10 +4399,13 @@ static int gfx_v9_0_get_cu_info(struct amdgpu_device *adev, { int i, j, k, counter, active_cu_number = 0; u32 mask, bitmap, ao_bitmap, ao_cu_mask = 0; + unsigned disable_masks[4 * 2]; if (!adev || !cu_info) return -EINVAL; + amdgpu_gfx_parse_disable_cu(disable_masks, 4, 2); + mutex_lock(&adev->grbm_idx_mutex); for (i = 0; i < adev->gfx.config.max_shader_engines; i++) { for (j = 0; j < adev->gfx.config.max_sh_per_se; j++) { @@ -4447,6 +4413,9 @@ static int gfx_v9_0_get_cu_info(struct amdgpu_device *adev, ao_bitmap = 0; counter = 0; gfx_v9_0_select_se_sh(adev, i, j, 0xffffffff); + if (i < 4 && j < 2) + gfx_v9_0_set_user_cu_inactive_bitmap( + adev, disable_masks[i * 2 + j]); bitmap = gfx_v9_0_get_cu_active_bitmap(adev); cu_info->bitmap[i][j] = bitmap; @@ -4459,7 +4428,9 @@ static int gfx_v9_0_get_cu_info(struct amdgpu_device *adev, mask <<= 1; } active_cu_number += counter; - ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + if (i < 2 && j < 2) + ao_cu_mask |= (ao_bitmap << (i * 16 + j * 8)); + cu_info->ao_cu_bitmap[i][j] = ao_bitmap; } } gfx_v9_0_select_se_sh(adev, 0xffffffff, 0xffffffff, 0xffffffff); diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c index ce68d609b619..d0214d942bfc 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v6_0.c @@ -794,14 +794,6 @@ static int gmc_v6_0_early_init(void *handle) gmc_v6_0_set_gart_funcs(adev); gmc_v6_0_set_irq_funcs(adev); - if (adev->flags & AMD_IS_APU) { - adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; - } else { - u32 tmp = RREG32(mmMC_SEQ_MISC0); - tmp &= MC_SEQ_MISC0__MT__MASK; - adev->mc.vram_type = gmc_v6_0_convert_vram_type(tmp); - } - return 0; } @@ -821,6 +813,14 @@ static int gmc_v6_0_sw_init(void *handle) int dma_bits; struct amdgpu_device *adev = (struct amdgpu_device *)handle; + if (adev->flags & AMD_IS_APU) { + adev->mc.vram_type = AMDGPU_VRAM_TYPE_UNKNOWN; + } else { + u32 tmp = RREG32(mmMC_SEQ_MISC0); + tmp &= MC_SEQ_MISC0__MT__MASK; + adev->mc.vram_type = gmc_v6_0_convert_vram_type(tmp); + } + r = amdgpu_irq_add_id(adev, AMDGPU_IH_CLIENTID_LEGACY, 146, &adev->mc.vm_fault); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 68172aace3ee..175ba5f9691c 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -695,6 +695,15 @@ static int gmc_v9_0_gart_enable(struct amdgpu_device *adev) else nbio_v6_1_hdp_flush(adev); + switch (adev->asic_type) { + case CHIP_RAVEN: + mmhub_v1_0_initialize_power_gating(adev); + mmhub_v1_0_update_power_gating(adev, true); + break; + default: + break; + } + r = gfxhub_v1_0_gart_enable(adev); if (r) return r; diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c index f50b5a77f45a..9804318f3488 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.c @@ -244,6 +244,224 @@ static void mmhub_v1_0_program_invalidation(struct amdgpu_device *adev) } } +struct pctl_data { + uint32_t index; + uint32_t data; +}; + +const struct pctl_data pctl0_data[] = { + {0x0, 0x7a640}, + {0x9, 0x2a64a}, + {0xd, 0x2a680}, + {0x11, 0x6a684}, + {0x19, 0xea68e}, + {0x29, 0xa69e}, + {0x2b, 0x34a6c0}, + {0x61, 0x83a707}, + {0xe6, 0x8a7a4}, + {0xf0, 0x1a7b8}, + {0xf3, 0xfa7cc}, + {0x104, 0x17a7dd}, + {0x11d, 0xa7dc}, + {0x11f, 0x12a7f5}, + {0x133, 0xa808}, + {0x135, 0x12a810}, + {0x149, 0x7a82c} +}; +#define PCTL0_DATA_LEN (sizeof(pctl0_data)/sizeof(pctl0_data[0])) + +#define PCTL0_RENG_EXEC_END_PTR 0x151 +#define PCTL0_STCTRL_REG_SAVE_RANGE0_BASE 0xa640 +#define PCTL0_STCTRL_REG_SAVE_RANGE0_LIMIT 0xa833 + +const struct pctl_data pctl1_data[] = { + {0x0, 0x39a000}, + {0x3b, 0x44a040}, + {0x81, 0x2a08d}, + {0x85, 0x6ba094}, + {0xf2, 0x18a100}, + {0x10c, 0x4a132}, + {0x112, 0xca141}, + {0x120, 0x2fa158}, + {0x151, 0x17a1d0}, + {0x16a, 0x1a1e9}, + {0x16d, 0x13a1ec}, + {0x182, 0x7a201}, + {0x18b, 0x3a20a}, + {0x190, 0x7a580}, + {0x199, 0xa590}, + {0x19b, 0x4a594}, + {0x1a1, 0x1a59c}, + {0x1a4, 0x7a82c}, + {0x1ad, 0xfa7cc}, + {0x1be, 0x17a7dd}, + {0x1d7, 0x12a810} +}; +#define PCTL1_DATA_LEN (sizeof(pctl1_data)/sizeof(pctl1_data[0])) + +#define PCTL1_RENG_EXEC_END_PTR 0x1ea +#define PCTL1_STCTRL_REG_SAVE_RANGE0_BASE 0xa000 +#define PCTL1_STCTRL_REG_SAVE_RANGE0_LIMIT 0xa20d +#define PCTL1_STCTRL_REG_SAVE_RANGE1_BASE 0xa580 +#define PCTL1_STCTRL_REG_SAVE_RANGE1_LIMIT 0xa59d +#define PCTL1_STCTRL_REG_SAVE_RANGE2_BASE 0xa82c +#define PCTL1_STCTRL_REG_SAVE_RANGE2_LIMIT 0xa833 + +static void mmhub_v1_0_power_gating_write_save_ranges(struct amdgpu_device *adev) +{ + uint32_t tmp = 0; + + /* PCTL0_STCTRL_REGISTER_SAVE_RANGE0 */ + tmp = REG_SET_FIELD(tmp, PCTL0_STCTRL_REGISTER_SAVE_RANGE0, + STCTRL_REGISTER_SAVE_BASE, + PCTL0_STCTRL_REG_SAVE_RANGE0_BASE); + tmp = REG_SET_FIELD(tmp, PCTL0_STCTRL_REGISTER_SAVE_RANGE0, + STCTRL_REGISTER_SAVE_LIMIT, + PCTL0_STCTRL_REG_SAVE_RANGE0_LIMIT); + WREG32_SOC15(MMHUB, 0, mmPCTL0_STCTRL_REGISTER_SAVE_RANGE0, tmp); + + /* PCTL1_STCTRL_REGISTER_SAVE_RANGE0 */ + tmp = 0; + tmp = REG_SET_FIELD(tmp, PCTL1_STCTRL_REGISTER_SAVE_RANGE0, + STCTRL_REGISTER_SAVE_BASE, + PCTL1_STCTRL_REG_SAVE_RANGE0_BASE); + tmp = REG_SET_FIELD(tmp, PCTL1_STCTRL_REGISTER_SAVE_RANGE0, + STCTRL_REGISTER_SAVE_LIMIT, + PCTL1_STCTRL_REG_SAVE_RANGE0_LIMIT); + WREG32_SOC15(MMHUB, 0, mmPCTL1_STCTRL_REGISTER_SAVE_RANGE0, tmp); + + /* PCTL1_STCTRL_REGISTER_SAVE_RANGE1 */ + tmp = 0; + tmp = REG_SET_FIELD(tmp, PCTL1_STCTRL_REGISTER_SAVE_RANGE1, + STCTRL_REGISTER_SAVE_BASE, + PCTL1_STCTRL_REG_SAVE_RANGE1_BASE); + tmp = REG_SET_FIELD(tmp, PCTL1_STCTRL_REGISTER_SAVE_RANGE1, + STCTRL_REGISTER_SAVE_LIMIT, + PCTL1_STCTRL_REG_SAVE_RANGE1_LIMIT); + WREG32_SOC15(MMHUB, 0, mmPCTL1_STCTRL_REGISTER_SAVE_RANGE1, tmp); + + /* PCTL1_STCTRL_REGISTER_SAVE_RANGE2 */ + tmp = 0; + tmp = REG_SET_FIELD(tmp, PCTL1_STCTRL_REGISTER_SAVE_RANGE2, + STCTRL_REGISTER_SAVE_BASE, + PCTL1_STCTRL_REG_SAVE_RANGE2_BASE); + tmp = REG_SET_FIELD(tmp, PCTL1_STCTRL_REGISTER_SAVE_RANGE2, + STCTRL_REGISTER_SAVE_LIMIT, + PCTL1_STCTRL_REG_SAVE_RANGE2_LIMIT); + WREG32_SOC15(MMHUB, 0, mmPCTL1_STCTRL_REGISTER_SAVE_RANGE2, tmp); +} + +void mmhub_v1_0_initialize_power_gating(struct amdgpu_device *adev) +{ + uint32_t pctl0_misc = 0; + uint32_t pctl0_reng_execute = 0; + uint32_t pctl1_misc = 0; + uint32_t pctl1_reng_execute = 0; + int i = 0; + + if (amdgpu_sriov_vf(adev)) + return; + + pctl0_misc = RREG32_SOC15(MMHUB, 0, mmPCTL0_MISC); + pctl0_reng_execute = RREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_EXECUTE); + pctl1_misc = RREG32_SOC15(MMHUB, 0, mmPCTL1_MISC); + pctl1_reng_execute = RREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE); + + /* Light sleep must be disabled before writing to pctl0 registers */ + pctl0_misc &= ~PCTL0_MISC__RENG_MEM_LS_ENABLE_MASK; + WREG32_SOC15(MMHUB, 0, mmPCTL0_MISC, pctl0_misc); + + /* Write data used to access ram of register engine */ + for (i = 0; i < PCTL0_DATA_LEN; i++) { + WREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_RAM_INDEX, + pctl0_data[i].index); + WREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_RAM_DATA, + pctl0_data[i].data); + } + + /* Set the reng execute end ptr for pctl0 */ + pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute, + PCTL0_RENG_EXECUTE, + RENG_EXECUTE_END_PTR, + PCTL0_RENG_EXEC_END_PTR); + WREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_EXECUTE, pctl0_reng_execute); + + /* Light sleep must be disabled before writing to pctl1 registers */ + pctl1_misc &= ~PCTL1_MISC__RENG_MEM_LS_ENABLE_MASK; + WREG32_SOC15(MMHUB, 0, mmPCTL1_MISC, pctl1_misc); + + /* Write data used to access ram of register engine */ + for (i = 0; i < PCTL1_DATA_LEN; i++) { + WREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_RAM_INDEX, + pctl1_data[i].index); + WREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_RAM_DATA, + pctl1_data[i].data); + } + + /* Set the reng execute end ptr for pctl1 */ + pctl1_reng_execute = REG_SET_FIELD(pctl1_reng_execute, + PCTL1_RENG_EXECUTE, + RENG_EXECUTE_END_PTR, + PCTL1_RENG_EXEC_END_PTR); + WREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE, pctl1_reng_execute); + + mmhub_v1_0_power_gating_write_save_ranges(adev); + + /* Re-enable light sleep */ + pctl0_misc |= PCTL0_MISC__RENG_MEM_LS_ENABLE_MASK; + WREG32_SOC15(MMHUB, 0, mmPCTL0_MISC, pctl0_misc); + pctl1_misc |= PCTL1_MISC__RENG_MEM_LS_ENABLE_MASK; + WREG32_SOC15(MMHUB, 0, mmPCTL1_MISC, pctl1_misc); +} + +void mmhub_v1_0_update_power_gating(struct amdgpu_device *adev, + bool enable) +{ + uint32_t pctl0_reng_execute = 0; + uint32_t pctl1_reng_execute = 0; + + if (amdgpu_sriov_vf(adev)) + return; + + pctl0_reng_execute = RREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_EXECUTE); + pctl1_reng_execute = RREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE); + + if (enable && adev->pg_flags & AMD_PG_SUPPORT_MMHUB) { + pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute, + PCTL0_RENG_EXECUTE, + RENG_EXECUTE_ON_PWR_UP, 1); + pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute, + PCTL0_RENG_EXECUTE, + RENG_EXECUTE_ON_REG_UPDATE, 1); + WREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_EXECUTE, pctl0_reng_execute); + + pctl1_reng_execute = REG_SET_FIELD(pctl1_reng_execute, + PCTL1_RENG_EXECUTE, + RENG_EXECUTE_ON_PWR_UP, 1); + pctl1_reng_execute = REG_SET_FIELD(pctl1_reng_execute, + PCTL1_RENG_EXECUTE, + RENG_EXECUTE_ON_REG_UPDATE, 1); + WREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE, pctl1_reng_execute); + + } else { + pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute, + PCTL0_RENG_EXECUTE, + RENG_EXECUTE_ON_PWR_UP, 0); + pctl0_reng_execute = REG_SET_FIELD(pctl0_reng_execute, + PCTL0_RENG_EXECUTE, + RENG_EXECUTE_ON_REG_UPDATE, 0); + WREG32_SOC15(MMHUB, 0, mmPCTL0_RENG_EXECUTE, pctl0_reng_execute); + + pctl1_reng_execute = REG_SET_FIELD(pctl1_reng_execute, + PCTL1_RENG_EXECUTE, + RENG_EXECUTE_ON_PWR_UP, 0); + pctl1_reng_execute = REG_SET_FIELD(pctl1_reng_execute, + PCTL1_RENG_EXECUTE, + RENG_EXECUTE_ON_REG_UPDATE, 0); + WREG32_SOC15(MMHUB, 0, mmPCTL1_RENG_EXECUTE, pctl1_reng_execute); + } +} + int mmhub_v1_0_gart_enable(struct amdgpu_device *adev) { if (amdgpu_sriov_vf(adev)) { diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.h b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.h index bbfacbcdc4a2..57bb940c0ecd 100644 --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.h +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v1_0.h @@ -32,6 +32,9 @@ void mmhub_v1_0_init(struct amdgpu_device *adev); int mmhub_v1_0_set_clockgating(struct amdgpu_device *adev, enum amd_clockgating_state state); void mmhub_v1_0_get_clockgating(struct amdgpu_device *adev, u32 *flags); +void mmhub_v1_0_initialize_power_gating(struct amdgpu_device *adev); +void mmhub_v1_0_update_power_gating(struct amdgpu_device *adev, + bool enable); extern const struct amd_ip_funcs mmhub_v1_0_ip_funcs; extern const struct amdgpu_ip_block_version mmhub_v1_0_ip_block; diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c index 20c1e539ff35..2258323a3c26 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v10_0.c @@ -96,8 +96,8 @@ int psp_v10_0_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cm header = (struct common_firmware_header *)ucode->fw; cmd->cmd_id = GFX_CMD_ID_LOAD_IP_FW; - cmd->cmd.cmd_load_ip_fw.fw_phy_addr_lo = (uint32_t)fw_mem_mc_addr; - cmd->cmd.cmd_load_ip_fw.fw_phy_addr_hi = (uint32_t)((uint64_t)fw_mem_mc_addr >> 32); + cmd->cmd.cmd_load_ip_fw.fw_phy_addr_lo = lower_32_bits(fw_mem_mc_addr); + cmd->cmd.cmd_load_ip_fw.fw_phy_addr_hi = upper_32_bits(fw_mem_mc_addr); cmd->cmd.cmd_load_ip_fw.fw_size = le32_to_cpu(header->ucode_size_bytes); ret = psp_v10_0_get_fw_type(ucode, &cmd->cmd.cmd_load_ip_fw.fw_type); @@ -172,10 +172,10 @@ int psp_v10_0_cmd_submit(struct psp_context *psp, write_frame = ring->ring_mem + (psp_write_ptr_reg / (sizeof(struct psp_gfx_rb_frame) / 4)); /* Update KM RB frame */ - write_frame->cmd_buf_addr_hi = (unsigned int)(cmd_buf_mc_addr >> 32); - write_frame->cmd_buf_addr_lo = (unsigned int)(cmd_buf_mc_addr); - write_frame->fence_addr_hi = (unsigned int)(fence_mc_addr >> 32); - write_frame->fence_addr_lo = (unsigned int)(fence_mc_addr); + write_frame->cmd_buf_addr_hi = upper_32_bits(cmd_buf_mc_addr); + write_frame->cmd_buf_addr_lo = lower_32_bits(cmd_buf_mc_addr); + write_frame->fence_addr_hi = upper_32_bits(fence_mc_addr); + write_frame->fence_addr_lo = lower_32_bits(fence_mc_addr); write_frame->fence_value = index; /* Update the write Pointer in DWORDs */ diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c index 6e5c6edabb84..c98d77d0c8f8 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c @@ -254,8 +254,8 @@ int psp_v3_1_prep_cmd_buf(struct amdgpu_firmware_info *ucode, struct psp_gfx_cmd memset(cmd, 0, sizeof(struct psp_gfx_cmd_resp)); cmd->cmd_id = GFX_CMD_ID_LOAD_IP_FW; - cmd->cmd.cmd_load_ip_fw.fw_phy_addr_lo = (uint32_t)fw_mem_mc_addr; - cmd->cmd.cmd_load_ip_fw.fw_phy_addr_hi = (uint32_t)((uint64_t)fw_mem_mc_addr >> 32); + cmd->cmd.cmd_load_ip_fw.fw_phy_addr_lo = lower_32_bits(fw_mem_mc_addr); + cmd->cmd.cmd_load_ip_fw.fw_phy_addr_hi = upper_32_bits(fw_mem_mc_addr); cmd->cmd.cmd_load_ip_fw.fw_size = ucode->ucode_size; ret = psp_v3_1_get_fw_type(ucode, &cmd->cmd.cmd_load_ip_fw.fw_type); @@ -375,10 +375,10 @@ int psp_v3_1_cmd_submit(struct psp_context *psp, memset(write_frame, 0, sizeof(struct psp_gfx_rb_frame)); /* Update KM RB frame */ - write_frame->cmd_buf_addr_hi = (unsigned int)(cmd_buf_mc_addr >> 32); - write_frame->cmd_buf_addr_lo = (unsigned int)(cmd_buf_mc_addr); - write_frame->fence_addr_hi = (unsigned int)(fence_mc_addr >> 32); - write_frame->fence_addr_lo = (unsigned int)(fence_mc_addr); + write_frame->cmd_buf_addr_hi = upper_32_bits(cmd_buf_mc_addr); + write_frame->cmd_buf_addr_lo = lower_32_bits(cmd_buf_mc_addr); + write_frame->fence_addr_hi = upper_32_bits(fence_mc_addr); + write_frame->fence_addr_lo = lower_32_bits(fence_mc_addr); write_frame->fence_value = index; /* Update the write Pointer in DWORDs */ diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index 5fdb05a0c88a..a7341d88a320 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -625,7 +625,8 @@ static int soc15_common_early_init(void *handle) AMD_CG_SUPPORT_MC_LS | AMD_CG_SUPPORT_SDMA_MGCG | AMD_CG_SUPPORT_SDMA_LS; - adev->pg_flags = AMD_PG_SUPPORT_SDMA; + adev->pg_flags = AMD_PG_SUPPORT_SDMA | + AMD_PG_SUPPORT_MMHUB; adev->external_rev_id = 0x1; break; default: diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c index 84d1ffd1eef9..035bbc98a63d 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c @@ -318,12 +318,13 @@ static struct kfd_process *create_process(const struct task_struct *thread) /* init process apertures*/ process->is_32bit_user_mode = in_compat_syscall(); - if (kfd_init_apertures(process) != 0) - goto err_init_apretures; + err = kfd_init_apertures(process); + if (err != 0) + goto err_init_apertures; return process; -err_init_apretures: +err_init_apertures: pqm_uninit(&process->pqm); err_process_pqm_init: hash_del_rcu(&process->kfd_processes); diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index beb2a81ab7da..70e8c20acb2f 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -184,6 +184,7 @@ enum amd_fan_ctrl_mode { #define AMD_PG_SUPPORT_SAMU (1 << 10) #define AMD_PG_SUPPORT_GFX_QUICK_MG (1 << 11) #define AMD_PG_SUPPORT_GFX_PIPELINE (1 << 12) +#define AMD_PG_SUPPORT_MMHUB (1 << 13) enum amd_pm_state_type { /* not used for dpm */ diff --git a/drivers/gpu/drm/amd/include/vi_structs.h b/drivers/gpu/drm/amd/include/vi_structs.h index b68f8efcdeae..ca93b5160ba6 100644 --- a/drivers/gpu/drm/amd/include/vi_structs.h +++ b/drivers/gpu/drm/amd/include/vi_structs.h @@ -195,6 +195,274 @@ struct vi_mqd { uint32_t compute_wave_restore_addr_lo; uint32_t compute_wave_restore_addr_hi; uint32_t compute_wave_restore_control; + uint32_t reserved9; + uint32_t reserved10; + uint32_t reserved11; + uint32_t reserved12; + uint32_t reserved13; + uint32_t reserved14; + uint32_t reserved15; + uint32_t reserved16; + uint32_t reserved17; + uint32_t reserved18; + uint32_t reserved19; + uint32_t reserved20; + uint32_t reserved21; + uint32_t reserved22; + uint32_t reserved23; + uint32_t reserved24; + uint32_t reserved25; + uint32_t reserved26; + uint32_t reserved27; + uint32_t reserved28; + uint32_t reserved29; + uint32_t reserved30; + uint32_t reserved31; + uint32_t reserved32; + uint32_t reserved33; + uint32_t reserved34; + uint32_t compute_user_data_0; + uint32_t compute_user_data_1; + uint32_t compute_user_data_2; + uint32_t compute_user_data_3; + uint32_t compute_user_data_4; + uint32_t compute_user_data_5; + uint32_t compute_user_data_6; + uint32_t compute_user_data_7; + uint32_t compute_user_data_8; + uint32_t compute_user_data_9; + uint32_t compute_user_data_10; + uint32_t compute_user_data_11; + uint32_t compute_user_data_12; + uint32_t compute_user_data_13; + uint32_t compute_user_data_14; + uint32_t compute_user_data_15; + uint32_t cp_compute_csinvoc_count_lo; + uint32_t cp_compute_csinvoc_count_hi; + uint32_t reserved35; + uint32_t reserved36; + uint32_t reserved37; + uint32_t cp_mqd_query_time_lo; + uint32_t cp_mqd_query_time_hi; + uint32_t cp_mqd_connect_start_time_lo; + uint32_t cp_mqd_connect_start_time_hi; + uint32_t cp_mqd_connect_end_time_lo; + uint32_t cp_mqd_connect_end_time_hi; + uint32_t cp_mqd_connect_end_wf_count; + uint32_t cp_mqd_connect_end_pq_rptr; + uint32_t cp_mqd_connect_endvi_sdma_mqd_pq_wptr; + uint32_t cp_mqd_connect_end_ib_rptr; + uint32_t reserved38; + uint32_t reserved39; + uint32_t cp_mqd_save_start_time_lo; + uint32_t cp_mqd_save_start_time_hi; + uint32_t cp_mqd_save_end_time_lo; + uint32_t cp_mqd_save_end_time_hi; + uint32_t cp_mqd_restore_start_time_lo; + uint32_t cp_mqd_restore_start_time_hi; + uint32_t cp_mqd_restore_end_time_lo; + uint32_t cp_mqd_restore_end_time_hi; + uint32_t disable_queue; + uint32_t reserved41; + uint32_t gds_cs_ctxsw_cnt0; + uint32_t gds_cs_ctxsw_cnt1; + uint32_t gds_cs_ctxsw_cnt2; + uint32_t gds_cs_ctxsw_cnt3; + uint32_t reserved42; + uint32_t reserved43; + uint32_t cp_pq_exe_status_lo; + uint32_t cp_pq_exe_status_hi; + uint32_t cp_packet_id_lo; + uint32_t cp_packet_id_hi; + uint32_t cp_packet_exe_status_lo; + uint32_t cp_packet_exe_status_hi; + uint32_t gds_save_base_addr_lo; + uint32_t gds_save_base_addr_hi; + uint32_t gds_save_mask_lo; + uint32_t gds_save_mask_hi; + uint32_t ctx_save_base_addr_lo; + uint32_t ctx_save_base_addr_hi; + uint32_t dynamic_cu_mask_addr_lo; + uint32_t dynamic_cu_mask_addr_hi; + uint32_t cp_mqd_base_addr_lo; + uint32_t cp_mqd_base_addr_hi; + uint32_t cp_hqd_active; + uint32_t cp_hqd_vmid; + uint32_t cp_hqd_persistent_state; + uint32_t cp_hqd_pipe_priority; + uint32_t cp_hqd_queue_priority; + uint32_t cp_hqd_quantum; + uint32_t cp_hqd_pq_base_lo; + uint32_t cp_hqd_pq_base_hi; + uint32_t cp_hqd_pq_rptr; + uint32_t cp_hqd_pq_rptr_report_addr_lo; + uint32_t cp_hqd_pq_rptr_report_addr_hi; + uint32_t cp_hqd_pq_wptr_poll_addr_lo; + uint32_t cp_hqd_pq_wptr_poll_addr_hi; + uint32_t cp_hqd_pq_doorbell_control; + uint32_t cp_hqd_pq_wptr; + uint32_t cp_hqd_pq_control; + uint32_t cp_hqd_ib_base_addr_lo; + uint32_t cp_hqd_ib_base_addr_hi; + uint32_t cp_hqd_ib_rptr; + uint32_t cp_hqd_ib_control; + uint32_t cp_hqd_iq_timer; + uint32_t cp_hqd_iq_rptr; + uint32_t cp_hqd_dequeue_request; + uint32_t cp_hqd_dma_offload; + uint32_t cp_hqd_sema_cmd; + uint32_t cp_hqd_msg_type; + uint32_t cp_hqd_atomic0_preop_lo; + uint32_t cp_hqd_atomic0_preop_hi; + uint32_t cp_hqd_atomic1_preop_lo; + uint32_t cp_hqd_atomic1_preop_hi; + uint32_t cp_hqd_hq_status0; + uint32_t cp_hqd_hq_control0; + uint32_t cp_mqd_control; + uint32_t cp_hqd_hq_status1; + uint32_t cp_hqd_hq_control1; + uint32_t cp_hqd_eop_base_addr_lo; + uint32_t cp_hqd_eop_base_addr_hi; + uint32_t cp_hqd_eop_control; + uint32_t cp_hqd_eop_rptr; + uint32_t cp_hqd_eop_wptr; + uint32_t cp_hqd_eop_done_events; + uint32_t cp_hqd_ctx_save_base_addr_lo; + uint32_t cp_hqd_ctx_save_base_addr_hi; + uint32_t cp_hqd_ctx_save_control; + uint32_t cp_hqd_cntl_stack_offset; + uint32_t cp_hqd_cntl_stack_size; + uint32_t cp_hqd_wg_state_offset; + uint32_t cp_hqd_ctx_save_size; + uint32_t cp_hqd_gds_resource_state; + uint32_t cp_hqd_error; + uint32_t cp_hqd_eop_wptr_mem; + uint32_t cp_hqd_eop_dones; + uint32_t reserved46; + uint32_t reserved47; + uint32_t reserved48; + uint32_t reserved49; + uint32_t reserved50; + uint32_t reserved51; + uint32_t reserved52; + uint32_t reserved53; + uint32_t reserved54; + uint32_t reserved55; + uint32_t iqtimer_pkt_header; + uint32_t iqtimer_pkt_dw0; + uint32_t iqtimer_pkt_dw1; + uint32_t iqtimer_pkt_dw2; + uint32_t iqtimer_pkt_dw3; + uint32_t iqtimer_pkt_dw4; + uint32_t iqtimer_pkt_dw5; + uint32_t iqtimer_pkt_dw6; + uint32_t iqtimer_pkt_dw7; + uint32_t iqtimer_pkt_dw8; + uint32_t iqtimer_pkt_dw9; + uint32_t iqtimer_pkt_dw10; + uint32_t iqtimer_pkt_dw11; + uint32_t iqtimer_pkt_dw12; + uint32_t iqtimer_pkt_dw13; + uint32_t iqtimer_pkt_dw14; + uint32_t iqtimer_pkt_dw15; + uint32_t iqtimer_pkt_dw16; + uint32_t iqtimer_pkt_dw17; + uint32_t iqtimer_pkt_dw18; + uint32_t iqtimer_pkt_dw19; + uint32_t iqtimer_pkt_dw20; + uint32_t iqtimer_pkt_dw21; + uint32_t iqtimer_pkt_dw22; + uint32_t iqtimer_pkt_dw23; + uint32_t iqtimer_pkt_dw24; + uint32_t iqtimer_pkt_dw25; + uint32_t iqtimer_pkt_dw26; + uint32_t iqtimer_pkt_dw27; + uint32_t iqtimer_pkt_dw28; + uint32_t iqtimer_pkt_dw29; + uint32_t iqtimer_pkt_dw30; + uint32_t iqtimer_pkt_dw31; + uint32_t reserved56; + uint32_t reserved57; + uint32_t reserved58; + uint32_t set_resources_header; + uint32_t set_resources_dw1; + uint32_t set_resources_dw2; + uint32_t set_resources_dw3; + uint32_t set_resources_dw4; + uint32_t set_resources_dw5; + uint32_t set_resources_dw6; + uint32_t set_resources_dw7; + uint32_t reserved59; + uint32_t reserved60; + uint32_t reserved61; + uint32_t reserved62; + uint32_t reserved63; + uint32_t reserved64; + uint32_t reserved65; + uint32_t reserved66; + uint32_t reserved67; + uint32_t reserved68; + uint32_t reserved69; + uint32_t reserved70; + uint32_t reserved71; + uint32_t reserved72; + uint32_t reserved73; + uint32_t reserved74; + uint32_t reserved75; + uint32_t reserved76; + uint32_t reserved77; + uint32_t reserved78; + uint32_t reserved_t[256]; +}; + +struct vi_mqd_allocation { + struct vi_mqd mqd; + uint32_t wptr_poll_mem; + uint32_t rptr_report_mem; + uint32_t dyamic_cu_mask; + uint32_t dyamic_rb_mask; +}; + +struct cz_mqd { + uint32_t header; + uint32_t compute_dispatch_initiator; + uint32_t compute_dim_x; + uint32_t compute_dim_y; + uint32_t compute_dim_z; + uint32_t compute_start_x; + uint32_t compute_start_y; + uint32_t compute_start_z; + uint32_t compute_num_thread_x; + uint32_t compute_num_thread_y; + uint32_t compute_num_thread_z; + uint32_t compute_pipelinestat_enable; + uint32_t compute_perfcount_enable; + uint32_t compute_pgm_lo; + uint32_t compute_pgm_hi; + uint32_t compute_tba_lo; + uint32_t compute_tba_hi; + uint32_t compute_tma_lo; + uint32_t compute_tma_hi; + uint32_t compute_pgm_rsrc1; + uint32_t compute_pgm_rsrc2; + uint32_t compute_vmid; + uint32_t compute_resource_limits; + uint32_t compute_static_thread_mgmt_se0; + uint32_t compute_static_thread_mgmt_se1; + uint32_t compute_tmpring_size; + uint32_t compute_static_thread_mgmt_se2; + uint32_t compute_static_thread_mgmt_se3; + uint32_t compute_restart_x; + uint32_t compute_restart_y; + uint32_t compute_restart_z; + uint32_t compute_thread_trace_enable; + uint32_t compute_misc_reserved; + uint32_t compute_dispatch_id; + uint32_t compute_threadgroup_id; + uint32_t compute_relaunch; + uint32_t compute_wave_restore_addr_lo; + uint32_t compute_wave_restore_addr_hi; + uint32_t compute_wave_restore_control; uint32_t reserved_39; uint32_t reserved_40; uint32_t reserved_41; diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c index f988ed204d9a..d6f097f44b6c 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_hwmgr.c @@ -2865,6 +2865,7 @@ static int vega10_get_pp_table_entry_callback_func(struct pp_hwmgr *hwmgr, void *state, struct pp_power_state *power_state, void *pp_table, uint32_t classification_flag) { + ATOM_Vega10_GFXCLK_Dependency_Record_V2 *patom_record_V2; struct vega10_power_state *vega10_power_state = cast_phw_vega10_power_state(&(power_state->hardware)); struct vega10_performance_level *performance_level; @@ -2941,11 +2942,16 @@ static int vega10_get_pp_table_entry_callback_func(struct pp_hwmgr *hwmgr, performance_level = &(vega10_power_state->performance_levels [vega10_power_state->performance_level_count++]); - performance_level->soc_clock = socclk_dep_table->entries - [state_entry->ucSocClockIndexHigh].ulClk; - performance_level->gfx_clock = gfxclk_dep_table->entries + [state_entry->ucSocClockIndexHigh].ulClk; + if (gfxclk_dep_table->ucRevId == 0) { + performance_level->gfx_clock = gfxclk_dep_table->entries [state_entry->ucGfxClockIndexHigh].ulClk; + } else if (gfxclk_dep_table->ucRevId == 1) { + patom_record_V2 = (ATOM_Vega10_GFXCLK_Dependency_Record_V2 *)gfxclk_dep_table->entries; + performance_level->gfx_clock = patom_record_V2[state_entry->ucGfxClockIndexHigh].ulClk; + } + performance_level->mem_clock = mclk_dep_table->entries [state_entry->ucMemClockIndexHigh].ulMemClk; return 0; @@ -3349,7 +3355,6 @@ static int vega10_populate_and_upload_sclk_mclk_dpm_levels( dpm_table-> gfx_table.dpm_levels[dpm_table->gfx_table.count - 1]. value = sclk; - if (phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, PHM_PlatformCaps_OD6PlusinACSupport) || phm_cap_enabled(hwmgr->platform_descriptor.platformCaps, @@ -3472,7 +3477,6 @@ static int vega10_populate_and_upload_sclk_mclk_dpm_levels( return result); } } - return result; } @@ -3828,13 +3832,18 @@ static int vega10_dpm_get_mclk(struct pp_hwmgr *hwmgr, bool low) static int vega10_get_gpu_power(struct pp_hwmgr *hwmgr, struct pp_gpu_power *query) { + uint32_t value; + PP_ASSERT_WITH_CODE(!smum_send_msg_to_smc(hwmgr->smumgr, PPSMC_MSG_GetCurrPkgPwr), "Failed to get current package power!", return -EINVAL); - return vega10_read_arg_from_smc(hwmgr->smumgr, - &query->average_gpu_power); + vega10_read_arg_from_smc(hwmgr->smumgr, &value); + /* power value is an integer */ + query->average_gpu_power = value << 8; + + return 0; } static int vega10_read_sensor(struct pp_hwmgr *hwmgr, int idx, diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_pptable.h b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_pptable.h index 52beea3bf6b7..b3e63003a789 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_pptable.h +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_pptable.h @@ -144,6 +144,15 @@ typedef struct _ATOM_Vega10_GFXCLK_Dependency_Record { USHORT usAVFSOffset; /* AVFS Voltage offset */ } ATOM_Vega10_GFXCLK_Dependency_Record; +typedef struct _ATOM_Vega10_GFXCLK_Dependency_Record_V2 { + ULONG ulClk; + UCHAR ucVddInd; + USHORT usCKSVOffsetandDisable; + USHORT usAVFSOffset; + UCHAR ucACGEnable; + UCHAR ucReserved[3]; +} ATOM_Vega10_GFXCLK_Dependency_Record_V2; + typedef struct _ATOM_Vega10_MCLK_Dependency_Record { ULONG ulMemClk; /* Clock Frequency */ UCHAR ucVddInd; /* SOC_VDD index */ diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c index 2b892e47d8dc..1623644ea49a 100644 --- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c +++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega10_processpptables.c @@ -585,6 +585,7 @@ static int get_gfxclk_voltage_dependency_table( uint32_t table_size, i; struct phm_ppt_v1_clock_voltage_dependency_table *clk_table; + ATOM_Vega10_GFXCLK_Dependency_Record_V2 *patom_record_v2; PP_ASSERT_WITH_CODE((clk_dep_table->ucNumEntries != 0), "Invalid PowerPlay Table!", return -1); @@ -601,18 +602,41 @@ static int get_gfxclk_voltage_dependency_table( clk_table->count = clk_dep_table->ucNumEntries; - for (i = 0; i < clk_table->count; i++) { - clk_table->entries[i].vddInd = + if (clk_dep_table->ucRevId == 0) { + for (i = 0; i < clk_table->count; i++) { + clk_table->entries[i].vddInd = clk_dep_table->entries[i].ucVddInd; - clk_table->entries[i].clk = + clk_table->entries[i].clk = le32_to_cpu(clk_dep_table->entries[i].ulClk); - clk_table->entries[i].cks_enable = - (((clk_dep_table->entries[i].usCKSVOffsetandDisable & 0x8000) + clk_table->entries[i].cks_enable = + (((le16_to_cpu(clk_dep_table->entries[i].usCKSVOffsetandDisable) & 0x8000) >> 15) == 0) ? 1 : 0; - clk_table->entries[i].cks_voffset = - (clk_dep_table->entries[i].usCKSVOffsetandDisable & 0x7F); - clk_table->entries[i].sclk_offset = - clk_dep_table->entries[i].usAVFSOffset; + clk_table->entries[i].cks_voffset = + le16_to_cpu(clk_dep_table->entries[i].usCKSVOffsetandDisable) & 0x7F; + clk_table->entries[i].sclk_offset = + le16_to_cpu(clk_dep_table->entries[i].usAVFSOffset); + } + } else if (clk_dep_table->ucRevId == 1) { + patom_record_v2 = (ATOM_Vega10_GFXCLK_Dependency_Record_V2 *)clk_dep_table->entries; + for (i = 0; i < clk_table->count; i++) { + clk_table->entries[i].vddInd = + patom_record_v2->ucVddInd; + clk_table->entries[i].clk = + le32_to_cpu(patom_record_v2->ulClk); + clk_table->entries[i].cks_enable = + (((le16_to_cpu(patom_record_v2->usCKSVOffsetandDisable) & 0x8000) + >> 15) == 0) ? 1 : 0; + clk_table->entries[i].cks_voffset = + le16_to_cpu(patom_record_v2->usCKSVOffsetandDisable) & 0x7F; + clk_table->entries[i].sclk_offset = + le16_to_cpu(patom_record_v2->usAVFSOffset); + patom_record_v2++; + } + } else { + kfree(clk_table); + PP_ASSERT_WITH_CODE(false, + "Unsupported GFXClockDependencyTable Revision!", + return -EINVAL); } *pp_vega10_clk_dep_table = clk_table; diff --git a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h index e07cab311c7a..b4af9e85dfa5 100644 --- a/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h +++ b/drivers/gpu/drm/amd/powerplay/inc/vega10_ppsmc.h @@ -124,8 +124,8 @@ typedef uint16_t PPSMC_Result; #define PPSMC_MSG_NumOfDisplays 0x56 #define PPSMC_MSG_ReadSerialNumTop32 0x58 #define PPSMC_MSG_ReadSerialNumBottom32 0x59 -#define PPSMC_MSG_GetCurrPkgPwr 0x5C -#define PPSMC_Message_Count 0x5D +#define PPSMC_MSG_GetCurrPkgPwr 0x61 +#define PPSMC_Message_Count 0x62 typedef int PPSMC_Msg; diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c index 39c7091866e8..652aaa43e95c 100644 --- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c +++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c @@ -72,7 +72,7 @@ static int cz_send_msg_to_smc_async(struct pp_smumgr *smumgr, result = SMUM_WAIT_FIELD_UNEQUAL(smumgr, SMU_MP1_SRBM2P_RESP_0, CONTENT, 0); if (result != 0) { - pr_err("cz_send_msg_to_smc_async failed\n"); + pr_err("cz_send_msg_to_smc_async (0x%04x) failed\n", msg); return result; } diff --git a/drivers/gpu/drm/armada/armada_fb.c b/drivers/gpu/drm/armada/armada_fb.c index 2a7eb6817c36..92e6b08ea64a 100644 --- a/drivers/gpu/drm/armada/armada_fb.c +++ b/drivers/gpu/drm/armada/armada_fb.c @@ -133,7 +133,7 @@ static struct drm_framebuffer *armada_fb_create(struct drm_device *dev, } /* Framebuffer objects must have a valid device address for scanout */ - if (obj->dev_addr == DMA_ERROR_CODE) { + if (!obj->mapped) { ret = -EINVAL; goto err_unref; } diff --git a/drivers/gpu/drm/armada/armada_gem.c b/drivers/gpu/drm/armada/armada_gem.c index d6c2a5d190eb..a76ca21d063b 100644 --- a/drivers/gpu/drm/armada/armada_gem.c +++ b/drivers/gpu/drm/armada/armada_gem.c @@ -175,6 +175,7 @@ armada_gem_linear_back(struct drm_device *dev, struct armada_gem_object *obj) obj->phys_addr = obj->linear->start; obj->dev_addr = obj->linear->start; + obj->mapped = true; } DRM_DEBUG_DRIVER("obj %p phys %#llx dev %#llx\n", obj, @@ -205,7 +206,6 @@ armada_gem_alloc_private_object(struct drm_device *dev, size_t size) return NULL; drm_gem_private_object_init(dev, &obj->obj, size); - obj->dev_addr = DMA_ERROR_CODE; DRM_DEBUG_DRIVER("alloc private obj %p size %zu\n", obj, size); @@ -229,8 +229,6 @@ static struct armada_gem_object *armada_gem_alloc_object(struct drm_device *dev, return NULL; } - obj->dev_addr = DMA_ERROR_CODE; - mapping = obj->obj.filp->f_mapping; mapping_set_gfp_mask(mapping, GFP_HIGHUSER | __GFP_RECLAIMABLE); @@ -610,5 +608,6 @@ int armada_gem_map_import(struct armada_gem_object *dobj) return -EINVAL; } dobj->dev_addr = sg_dma_address(dobj->sgt->sgl); + dobj->mapped = true; return 0; } diff --git a/drivers/gpu/drm/armada/armada_gem.h b/drivers/gpu/drm/armada/armada_gem.h index b88d2b9853c7..6e524e0676bb 100644 --- a/drivers/gpu/drm/armada/armada_gem.h +++ b/drivers/gpu/drm/armada/armada_gem.h @@ -16,6 +16,7 @@ struct armada_gem_object { void *addr; phys_addr_t phys_addr; resource_size_t dev_addr; + bool mapped; struct drm_mm_node *linear; /* for linear backed */ struct page *page; /* for page backed */ struct sg_table *sgt; /* for imported */ diff --git a/drivers/gpu/drm/bridge/adv7511/adv7511_audio.c b/drivers/gpu/drm/bridge/adv7511/adv7511_audio.c index cf92ebfe6ab7..67469c26bae8 100644 --- a/drivers/gpu/drm/bridge/adv7511/adv7511_audio.c +++ b/drivers/gpu/drm/bridge/adv7511/adv7511_audio.c @@ -11,6 +11,7 @@ #include <sound/hdmi-codec.h> #include <sound/pcm.h> #include <sound/soc.h> +#include <linux/of_graph.h> #include "adv7511.h" @@ -182,10 +183,31 @@ static void audio_shutdown(struct device *dev, void *data) { } +static int adv7511_hdmi_i2s_get_dai_id(struct snd_soc_component *component, + struct device_node *endpoint) +{ + struct of_endpoint of_ep; + int ret; + + ret = of_graph_parse_endpoint(endpoint, &of_ep); + if (ret < 0) + return ret; + + /* + * HDMI sound should be located as reg = <2> + * Then, it is sound port 0 + */ + if (of_ep.port == 2) + return 0; + + return -EINVAL; +} + static const struct hdmi_codec_ops adv7511_codec_ops = { .hw_params = adv7511_hdmi_hw_params, .audio_shutdown = audio_shutdown, .audio_startup = audio_startup, + .get_dai_id = adv7511_hdmi_i2s_get_dai_id, }; static struct hdmi_codec_pdata codec_data = { diff --git a/drivers/gpu/drm/bridge/panel.c b/drivers/gpu/drm/bridge/panel.c index 65ab28cc2946..685c1a480201 100644 --- a/drivers/gpu/drm/bridge/panel.c +++ b/drivers/gpu/drm/bridge/panel.c @@ -160,7 +160,7 @@ struct drm_bridge *drm_panel_bridge_add(struct drm_panel *panel, struct panel_bridge *panel_bridge; if (!panel) - return ERR_PTR(EINVAL); + return ERR_PTR(-EINVAL); panel_bridge = devm_kzalloc(panel->dev, sizeof(*panel_bridge), GFP_KERNEL); diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-i2s-audio.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-i2s-audio.c index aaf287d2e91d..b2cf59f54c88 100644 --- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi-i2s-audio.c +++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi-i2s-audio.c @@ -82,9 +82,30 @@ static void dw_hdmi_i2s_audio_shutdown(struct device *dev, void *data) hdmi_write(audio, HDMI_AUD_CONF0_SW_RESET, HDMI_AUD_CONF0); } +static int dw_hdmi_i2s_get_dai_id(struct snd_soc_component *component, + struct device_node *endpoint) +{ + struct of_endpoint of_ep; + int ret; + + ret = of_graph_parse_endpoint(endpoint, &of_ep); + if (ret < 0) + return ret; + + /* + * HDMI sound should be located as reg = <2> + * Then, it is sound port 0 + */ + if (of_ep.port == 2) + return 0; + + return -EINVAL; +} + static struct hdmi_codec_ops dw_hdmi_i2s_ops = { .hw_params = dw_hdmi_i2s_hw_params, .audio_shutdown = dw_hdmi_i2s_audio_shutdown, + .get_dai_id = dw_hdmi_i2s_get_dai_id, }; static int snd_dw_hdmi_probe(struct platform_device *pdev) diff --git a/drivers/gpu/drm/drm_bufs.c b/drivers/gpu/drm/drm_bufs.c index adb1dd7fde5f..1ee84dd802d4 100644 --- a/drivers/gpu/drm/drm_bufs.c +++ b/drivers/gpu/drm/drm_bufs.c @@ -1258,11 +1258,11 @@ int drm_legacy_addbufs(struct drm_device *dev, void *data, * lock, preventing of allocating more buffers after this call. Information * about each requested buffer is then copied into user space. */ -int drm_legacy_infobufs(struct drm_device *dev, void *data, - struct drm_file *file_priv) +int __drm_legacy_infobufs(struct drm_device *dev, + void *data, int *p, + int (*f)(void *, int, struct drm_buf_entry *)) { struct drm_device_dma *dma = dev->dma; - struct drm_buf_info *request = data; int i; int count; @@ -1290,26 +1290,12 @@ int drm_legacy_infobufs(struct drm_device *dev, void *data, DRM_DEBUG("count = %d\n", count); - if (request->count >= count) { + if (*p >= count) { for (i = 0, count = 0; i < DRM_MAX_ORDER + 1; i++) { - if (dma->bufs[i].buf_count) { - struct drm_buf_desc __user *to = - &request->list[count]; - struct drm_buf_entry *from = &dma->bufs[i]; - if (copy_to_user(&to->count, - &from->buf_count, - sizeof(from->buf_count)) || - copy_to_user(&to->size, - &from->buf_size, - sizeof(from->buf_size)) || - copy_to_user(&to->low_mark, - &from->low_mark, - sizeof(from->low_mark)) || - copy_to_user(&to->high_mark, - &from->high_mark, - sizeof(from->high_mark))) + struct drm_buf_entry *from = &dma->bufs[i]; + if (from->buf_count) { + if (f(data, count, from) < 0) return -EFAULT; - DRM_DEBUG("%d %d %d %d %d\n", i, dma->bufs[i].buf_count, @@ -1320,11 +1306,29 @@ int drm_legacy_infobufs(struct drm_device *dev, void *data, } } } - request->count = count; + *p = count; return 0; } +static int copy_one_buf(void *data, int count, struct drm_buf_entry *from) +{ + struct drm_buf_info *request = data; + struct drm_buf_desc __user *to = &request->list[count]; + struct drm_buf_desc v = {.count = from->buf_count, + .size = from->buf_size, + .low_mark = from->low_mark, + .high_mark = from->high_mark}; + return copy_to_user(to, &v, offsetof(struct drm_buf_desc, flags)); +} + +int drm_legacy_infobufs(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + struct drm_buf_info *request = data; + return __drm_legacy_infobufs(dev, data, &request->count, copy_one_buf); +} + /** * Specifies a low and high water mark for buffer allocation * @@ -1439,15 +1443,15 @@ int drm_legacy_freebufs(struct drm_device *dev, void *data, * offset equal to 0, which drm_mmap() interpretes as PCI buffers and calls * drm_mmap_dma(). */ -int drm_legacy_mapbufs(struct drm_device *dev, void *data, - struct drm_file *file_priv) +int __drm_legacy_mapbufs(struct drm_device *dev, void *data, int *p, + void __user **v, + int (*f)(void *, int, unsigned long, + struct drm_buf *), + struct drm_file *file_priv) { struct drm_device_dma *dma = dev->dma; int retcode = 0; - const int zero = 0; unsigned long virtual; - unsigned long address; - struct drm_buf_map *request = data; int i; if (!drm_core_check_feature(dev, DRIVER_LEGACY)) @@ -1467,7 +1471,7 @@ int drm_legacy_mapbufs(struct drm_device *dev, void *data, dev->buf_use++; /* Can't allocate more after this call */ spin_unlock(&dev->buf_lock); - if (request->count >= dma->buf_count) { + if (*p >= dma->buf_count) { if ((dev->agp && (dma->flags & _DRM_DMA_USE_AGP)) || (drm_core_check_feature(dev, DRIVER_SG) && (dma->flags & _DRM_DMA_USE_SG))) { @@ -1492,41 +1496,51 @@ int drm_legacy_mapbufs(struct drm_device *dev, void *data, retcode = (signed long)virtual; goto done; } - request->virtual = (void __user *)virtual; + *v = (void __user *)virtual; for (i = 0; i < dma->buf_count; i++) { - if (copy_to_user(&request->list[i].idx, - &dma->buflist[i]->idx, - sizeof(request->list[0].idx))) { - retcode = -EFAULT; - goto done; - } - if (copy_to_user(&request->list[i].total, - &dma->buflist[i]->total, - sizeof(request->list[0].total))) { - retcode = -EFAULT; - goto done; - } - if (copy_to_user(&request->list[i].used, - &zero, sizeof(zero))) { - retcode = -EFAULT; - goto done; - } - address = virtual + dma->buflist[i]->offset; /* *** */ - if (copy_to_user(&request->list[i].address, - &address, sizeof(address))) { + if (f(data, i, virtual, dma->buflist[i]) < 0) { retcode = -EFAULT; goto done; } } } done: - request->count = dma->buf_count; - DRM_DEBUG("%d buffers, retcode = %d\n", request->count, retcode); + *p = dma->buf_count; + DRM_DEBUG("%d buffers, retcode = %d\n", *p, retcode); return retcode; } +static int map_one_buf(void *data, int idx, unsigned long virtual, + struct drm_buf *buf) +{ + struct drm_buf_map *request = data; + unsigned long address = virtual + buf->offset; /* *** */ + + if (copy_to_user(&request->list[idx].idx, &buf->idx, + sizeof(request->list[0].idx))) + return -EFAULT; + if (copy_to_user(&request->list[idx].total, &buf->total, + sizeof(request->list[0].total))) + return -EFAULT; + if (clear_user(&request->list[idx].used, sizeof(int))) + return -EFAULT; + if (copy_to_user(&request->list[idx].address, &address, + sizeof(address))) + return -EFAULT; + return 0; +} + +int drm_legacy_mapbufs(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + struct drm_buf_map *request = data; + return __drm_legacy_mapbufs(dev, data, &request->count, + &request->virtual, map_one_buf, + file_priv); +} + int drm_legacy_dma_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv) { diff --git a/drivers/gpu/drm/drm_dp_aux_dev.c b/drivers/gpu/drm/drm_dp_aux_dev.c index ec1ed94b2390..d34e5096887a 100644 --- a/drivers/gpu/drm/drm_dp_aux_dev.c +++ b/drivers/gpu/drm/drm_dp_aux_dev.c @@ -32,6 +32,7 @@ #include <linux/kernel.h> #include <linux/module.h> #include <linux/uaccess.h> +#include <linux/uio.h> #include <drm/drm_dp_helper.h> #include <drm/drm_crtc.h> #include <drm/drmP.h> @@ -140,101 +141,83 @@ static loff_t auxdev_llseek(struct file *file, loff_t offset, int whence) return fixed_size_llseek(file, offset, whence, AUX_MAX_OFFSET); } -static ssize_t auxdev_read(struct file *file, char __user *buf, size_t count, - loff_t *offset) +static ssize_t auxdev_read_iter(struct kiocb *iocb, struct iov_iter *to) { - size_t bytes_pending, num_bytes_processed = 0; - struct drm_dp_aux_dev *aux_dev = file->private_data; + struct drm_dp_aux_dev *aux_dev = iocb->ki_filp->private_data; + loff_t pos = iocb->ki_pos; ssize_t res = 0; if (!atomic_inc_not_zero(&aux_dev->usecount)) return -ENODEV; - bytes_pending = min((loff_t)count, AUX_MAX_OFFSET - (*offset)); - - if (!access_ok(VERIFY_WRITE, buf, bytes_pending)) { - res = -EFAULT; - goto out; - } + iov_iter_truncate(to, AUX_MAX_OFFSET - pos); - while (bytes_pending > 0) { - uint8_t localbuf[DP_AUX_MAX_PAYLOAD_BYTES]; - ssize_t todo = min_t(size_t, bytes_pending, sizeof(localbuf)); + while (iov_iter_count(to)) { + uint8_t buf[DP_AUX_MAX_PAYLOAD_BYTES]; + ssize_t todo = min(iov_iter_count(to), sizeof(buf)); if (signal_pending(current)) { - res = num_bytes_processed ? - num_bytes_processed : -ERESTARTSYS; - goto out; + res = -ERESTARTSYS; + break; } - res = drm_dp_dpcd_read(aux_dev->aux, *offset, localbuf, todo); - if (res <= 0) { - res = num_bytes_processed ? num_bytes_processed : res; - goto out; - } - if (__copy_to_user(buf + num_bytes_processed, localbuf, res)) { - res = num_bytes_processed ? - num_bytes_processed : -EFAULT; - goto out; + res = drm_dp_dpcd_read(aux_dev->aux, pos, buf, todo); + if (res <= 0) + break; + + if (copy_to_iter(buf, res, to) != res) { + res = -EFAULT; + break; } - bytes_pending -= res; - *offset += res; - num_bytes_processed += res; - res = num_bytes_processed; + + pos += res; } -out: + if (pos != iocb->ki_pos) + res = pos - iocb->ki_pos; + iocb->ki_pos = pos; + atomic_dec(&aux_dev->usecount); wake_up_atomic_t(&aux_dev->usecount); return res; } -static ssize_t auxdev_write(struct file *file, const char __user *buf, - size_t count, loff_t *offset) +static ssize_t auxdev_write_iter(struct kiocb *iocb, struct iov_iter *from) { - size_t bytes_pending, num_bytes_processed = 0; - struct drm_dp_aux_dev *aux_dev = file->private_data; + struct drm_dp_aux_dev *aux_dev = iocb->ki_filp->private_data; + loff_t pos = iocb->ki_pos; ssize_t res = 0; if (!atomic_inc_not_zero(&aux_dev->usecount)) return -ENODEV; - bytes_pending = min((loff_t)count, AUX_MAX_OFFSET - *offset); - - if (!access_ok(VERIFY_READ, buf, bytes_pending)) { - res = -EFAULT; - goto out; - } + iov_iter_truncate(from, AUX_MAX_OFFSET - pos); - while (bytes_pending > 0) { - uint8_t localbuf[DP_AUX_MAX_PAYLOAD_BYTES]; - ssize_t todo = min_t(size_t, bytes_pending, sizeof(localbuf)); + while (iov_iter_count(from)) { + uint8_t buf[DP_AUX_MAX_PAYLOAD_BYTES]; + ssize_t todo = min(iov_iter_count(from), sizeof(buf)); if (signal_pending(current)) { - res = num_bytes_processed ? - num_bytes_processed : -ERESTARTSYS; - goto out; + res = -ERESTARTSYS; + break; } - if (__copy_from_user(localbuf, - buf + num_bytes_processed, todo)) { - res = num_bytes_processed ? - num_bytes_processed : -EFAULT; - goto out; + if (!copy_from_iter_full(buf, todo, from)) { + res = -EFAULT; + break; } - res = drm_dp_dpcd_write(aux_dev->aux, *offset, localbuf, todo); - if (res <= 0) { - res = num_bytes_processed ? num_bytes_processed : res; - goto out; - } - bytes_pending -= res; - *offset += res; - num_bytes_processed += res; - res = num_bytes_processed; + res = drm_dp_dpcd_write(aux_dev->aux, pos, buf, todo); + if (res <= 0) + break; + + pos += res; } -out: + if (pos != iocb->ki_pos) + res = pos - iocb->ki_pos; + iocb->ki_pos = pos; + atomic_dec(&aux_dev->usecount); wake_up_atomic_t(&aux_dev->usecount); return res; @@ -251,8 +234,8 @@ static int auxdev_release(struct inode *inode, struct file *file) static const struct file_operations auxdev_fops = { .owner = THIS_MODULE, .llseek = auxdev_llseek, - .read = auxdev_read, - .write = auxdev_write, + .read_iter = auxdev_read_iter, + .write_iter = auxdev_write_iter, .open = auxdev_open, .release = auxdev_release, }; diff --git a/drivers/gpu/drm/drm_framebuffer.c b/drivers/gpu/drm/drm_framebuffer.c index 607ef3a97c42..af279844d7ce 100644 --- a/drivers/gpu/drm/drm_framebuffer.c +++ b/drivers/gpu/drm/drm_framebuffer.c @@ -832,6 +832,7 @@ unlock: drm_atomic_clean_old_fb(dev, plane_mask, ret); if (ret == -EDEADLK) { + drm_atomic_state_clear(state); drm_modeset_backoff(&ctx); goto retry; } diff --git a/drivers/gpu/drm/drm_internal.h b/drivers/gpu/drm/drm_internal.h index 068b685608cf..4e906b82a170 100644 --- a/drivers/gpu/drm/drm_internal.h +++ b/drivers/gpu/drm/drm_internal.h @@ -152,6 +152,10 @@ static inline int drm_debugfs_crtc_crc_add(struct drm_crtc *crtc) #endif +drm_ioctl_t drm_version; +drm_ioctl_t drm_getunique; +drm_ioctl_t drm_getclient; + /* drm_syncobj.c */ void drm_syncobj_open(struct drm_file *file_private); void drm_syncobj_release(struct drm_file *file_private); diff --git a/drivers/gpu/drm/drm_ioc32.c b/drivers/gpu/drm/drm_ioc32.c index ae386783e3ea..f8e96e648acf 100644 --- a/drivers/gpu/drm/drm_ioc32.c +++ b/drivers/gpu/drm/drm_ioc32.c @@ -32,6 +32,9 @@ #include <linux/export.h> #include <drm/drmP.h> +#include "drm_legacy.h" +#include "drm_internal.h" +#include "drm_crtc_internal.h" #define DRM_IOCTL_VERSION32 DRM_IOWR(0x00, drm_version32_t) #define DRM_IOCTL_GET_UNIQUE32 DRM_IOWR(0x01, drm_unique32_t) @@ -87,39 +90,31 @@ static int compat_drm_version(struct file *file, unsigned int cmd, unsigned long arg) { drm_version32_t v32; - struct drm_version __user *version; + struct drm_version v; int err; if (copy_from_user(&v32, (void __user *)arg, sizeof(v32))) return -EFAULT; - version = compat_alloc_user_space(sizeof(*version)); - if (!version) - return -EFAULT; - if (__put_user(v32.name_len, &version->name_len) - || __put_user((void __user *)(unsigned long)v32.name, - &version->name) - || __put_user(v32.date_len, &version->date_len) - || __put_user((void __user *)(unsigned long)v32.date, - &version->date) - || __put_user(v32.desc_len, &version->desc_len) - || __put_user((void __user *)(unsigned long)v32.desc, - &version->desc)) - return -EFAULT; - - err = drm_ioctl(file, - DRM_IOCTL_VERSION, (unsigned long)version); + v = (struct drm_version) { + .name_len = v32.name_len, + .name = compat_ptr(v32.name), + .date_len = v32.date_len, + .date = compat_ptr(v32.date), + .desc_len = v32.desc_len, + .desc = compat_ptr(v32.desc), + }; + err = drm_ioctl_kernel(file, drm_version, &v, + DRM_UNLOCKED|DRM_RENDER_ALLOW|DRM_CONTROL_ALLOW); if (err) return err; - if (__get_user(v32.version_major, &version->version_major) - || __get_user(v32.version_minor, &version->version_minor) - || __get_user(v32.version_patchlevel, &version->version_patchlevel) - || __get_user(v32.name_len, &version->name_len) - || __get_user(v32.date_len, &version->date_len) - || __get_user(v32.desc_len, &version->desc_len)) - return -EFAULT; - + v32.version_major = v.version_major; + v32.version_minor = v.version_minor; + v32.version_patchlevel = v.version_patchlevel; + v32.name_len = v.name_len; + v32.date_len = v.date_len; + v32.desc_len = v.desc_len; if (copy_to_user((void __user *)arg, &v32, sizeof(v32))) return -EFAULT; return 0; @@ -134,26 +129,21 @@ static int compat_drm_getunique(struct file *file, unsigned int cmd, unsigned long arg) { drm_unique32_t uq32; - struct drm_unique __user *u; + struct drm_unique uq; int err; if (copy_from_user(&uq32, (void __user *)arg, sizeof(uq32))) return -EFAULT; + uq = (struct drm_unique){ + .unique_len = uq32.unique_len, + .unique = compat_ptr(uq32.unique), + }; - u = compat_alloc_user_space(sizeof(*u)); - if (!u) - return -EFAULT; - if (__put_user(uq32.unique_len, &u->unique_len) - || __put_user((void __user *)(unsigned long)uq32.unique, - &u->unique)) - return -EFAULT; - - err = drm_ioctl(file, DRM_IOCTL_GET_UNIQUE, (unsigned long)u); + err = drm_ioctl_kernel(file, drm_getunique, &uq, DRM_UNLOCKED); if (err) return err; - if (__get_user(uq32.unique_len, &u->unique_len)) - return -EFAULT; + uq32.unique_len = uq.unique_len; if (copy_to_user((void __user *)arg, &uq32, sizeof(uq32))) return -EFAULT; return 0; @@ -162,21 +152,8 @@ static int compat_drm_getunique(struct file *file, unsigned int cmd, static int compat_drm_setunique(struct file *file, unsigned int cmd, unsigned long arg) { - drm_unique32_t uq32; - struct drm_unique __user *u; - - if (copy_from_user(&uq32, (void __user *)arg, sizeof(uq32))) - return -EFAULT; - - u = compat_alloc_user_space(sizeof(*u)); - if (!u) - return -EFAULT; - if (__put_user(uq32.unique_len, &u->unique_len) - || __put_user((void __user *)(unsigned long)uq32.unique, - &u->unique)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_SET_UNIQUE, (unsigned long)u); + /* it's dead */ + return -EINVAL; } typedef struct drm_map32 { @@ -193,32 +170,23 @@ static int compat_drm_getmap(struct file *file, unsigned int cmd, { drm_map32_t __user *argp = (void __user *)arg; drm_map32_t m32; - struct drm_map __user *map; - int idx, err; - void *handle; - - if (get_user(idx, &argp->offset)) - return -EFAULT; + struct drm_map map; + int err; - map = compat_alloc_user_space(sizeof(*map)); - if (!map) - return -EFAULT; - if (__put_user(idx, &map->offset)) + if (copy_from_user(&m32, argp, sizeof(m32))) return -EFAULT; - err = drm_ioctl(file, DRM_IOCTL_GET_MAP, (unsigned long)map); + map.offset = m32.offset; + err = drm_ioctl_kernel(file, drm_legacy_getmap_ioctl, &map, DRM_UNLOCKED); if (err) return err; - if (__get_user(m32.offset, &map->offset) - || __get_user(m32.size, &map->size) - || __get_user(m32.type, &map->type) - || __get_user(m32.flags, &map->flags) - || __get_user(handle, &map->handle) - || __get_user(m32.mtrr, &map->mtrr)) - return -EFAULT; - - m32.handle = (unsigned long)handle; + m32.offset = map.offset; + m32.size = map.size; + m32.type = map.type; + m32.flags = map.flags; + m32.handle = ptr_to_compat(map.handle); + m32.mtrr = map.mtrr; if (copy_to_user(argp, &m32, sizeof(m32))) return -EFAULT; return 0; @@ -230,35 +198,28 @@ static int compat_drm_addmap(struct file *file, unsigned int cmd, { drm_map32_t __user *argp = (void __user *)arg; drm_map32_t m32; - struct drm_map __user *map; + struct drm_map map; int err; - void *handle; if (copy_from_user(&m32, argp, sizeof(m32))) return -EFAULT; - map = compat_alloc_user_space(sizeof(*map)); - if (!map) - return -EFAULT; - if (__put_user(m32.offset, &map->offset) - || __put_user(m32.size, &map->size) - || __put_user(m32.type, &map->type) - || __put_user(m32.flags, &map->flags)) - return -EFAULT; + map.offset = m32.offset; + map.size = m32.size; + map.type = m32.type; + map.flags = m32.flags; - err = drm_ioctl(file, DRM_IOCTL_ADD_MAP, (unsigned long)map); + err = drm_ioctl_kernel(file, drm_legacy_addmap_ioctl, &map, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); if (err) return err; - if (__get_user(m32.offset, &map->offset) - || __get_user(m32.mtrr, &map->mtrr) - || __get_user(handle, &map->handle)) - return -EFAULT; - - m32.handle = (unsigned long)handle; - if (m32.handle != (unsigned long)handle) + m32.offset = map.offset; + m32.mtrr = map.mtrr; + m32.handle = ptr_to_compat(map.handle); + if (map.handle != compat_ptr(m32.handle)) pr_err_ratelimited("compat_drm_addmap truncated handle %p for type %d offset %x\n", - handle, m32.type, m32.offset); + map.handle, m32.type, m32.offset); if (copy_to_user(argp, &m32, sizeof(m32))) return -EFAULT; @@ -270,19 +231,13 @@ static int compat_drm_rmmap(struct file *file, unsigned int cmd, unsigned long arg) { drm_map32_t __user *argp = (void __user *)arg; - struct drm_map __user *map; + struct drm_map map; u32 handle; if (get_user(handle, &argp->handle)) return -EFAULT; - - map = compat_alloc_user_space(sizeof(*map)); - if (!map) - return -EFAULT; - if (__put_user((void *)(unsigned long)handle, &map->handle)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RM_MAP, (unsigned long)map); + map.handle = compat_ptr(handle); + return drm_ioctl_kernel(file, drm_legacy_rmmap_ioctl, &map, DRM_AUTH); } typedef struct drm_client32 { @@ -299,29 +254,24 @@ static int compat_drm_getclient(struct file *file, unsigned int cmd, { drm_client32_t c32; drm_client32_t __user *argp = (void __user *)arg; - struct drm_client __user *client; - int idx, err; + struct drm_client client; + int err; - if (get_user(idx, &argp->idx)) + if (copy_from_user(&c32, argp, sizeof(c32))) return -EFAULT; - client = compat_alloc_user_space(sizeof(*client)); - if (!client) - return -EFAULT; - if (__put_user(idx, &client->idx)) - return -EFAULT; + client.idx = c32.idx; - err = drm_ioctl(file, DRM_IOCTL_GET_CLIENT, (unsigned long)client); + err = drm_ioctl_kernel(file, drm_getclient, &client, DRM_UNLOCKED); if (err) return err; - if (__get_user(c32.idx, &client->idx) - || __get_user(c32.auth, &client->auth) - || __get_user(c32.pid, &client->pid) - || __get_user(c32.uid, &client->uid) - || __get_user(c32.magic, &client->magic) - || __get_user(c32.iocs, &client->iocs)) - return -EFAULT; + c32.idx = client.idx; + c32.auth = client.auth; + c32.pid = client.pid; + c32.uid = client.uid; + c32.magic = client.magic; + c32.iocs = client.iocs; if (copy_to_user(argp, &c32, sizeof(c32))) return -EFAULT; @@ -339,28 +289,14 @@ typedef struct drm_stats32 { static int compat_drm_getstats(struct file *file, unsigned int cmd, unsigned long arg) { - drm_stats32_t s32; drm_stats32_t __user *argp = (void __user *)arg; - struct drm_stats __user *stats; - int i, err; - - memset(&s32, 0, sizeof(drm_stats32_t)); - stats = compat_alloc_user_space(sizeof(*stats)); - if (!stats) - return -EFAULT; + int err; - err = drm_ioctl(file, DRM_IOCTL_GET_STATS, (unsigned long)stats); + err = drm_ioctl_kernel(file, drm_noop, NULL, DRM_UNLOCKED); if (err) return err; - if (__get_user(s32.count, &stats->count)) - return -EFAULT; - for (i = 0; i < 15; ++i) - if (__get_user(s32.data[i].value, &stats->data[i].value) - || __get_user(s32.data[i].type, &stats->data[i].type)) - return -EFAULT; - - if (copy_to_user(argp, &s32, sizeof(s32))) + if (clear_user(argp, sizeof(drm_stats32_t))) return -EFAULT; return 0; } @@ -378,26 +314,28 @@ static int compat_drm_addbufs(struct file *file, unsigned int cmd, unsigned long arg) { drm_buf_desc32_t __user *argp = (void __user *)arg; - struct drm_buf_desc __user *buf; + drm_buf_desc32_t desc32; + struct drm_buf_desc desc; int err; - unsigned long agp_start; - buf = compat_alloc_user_space(sizeof(*buf)); - if (!buf || !access_ok(VERIFY_WRITE, argp, sizeof(*argp))) + if (copy_from_user(&desc32, argp, sizeof(drm_buf_desc32_t))) return -EFAULT; - if (__copy_in_user(buf, argp, offsetof(drm_buf_desc32_t, agp_start)) - || __get_user(agp_start, &argp->agp_start) - || __put_user(agp_start, &buf->agp_start)) - return -EFAULT; + desc = (struct drm_buf_desc){ + desc32.count, desc32.size, desc32.low_mark, desc32.high_mark, + desc32.flags, desc32.agp_start + }; - err = drm_ioctl(file, DRM_IOCTL_ADD_BUFS, (unsigned long)buf); + err = drm_ioctl_kernel(file, drm_legacy_addbufs, &desc, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); if (err) return err; - if (__copy_in_user(argp, buf, offsetof(drm_buf_desc32_t, agp_start)) - || __get_user(agp_start, &buf->agp_start) - || __put_user(agp_start, &argp->agp_start)) + desc32 = (drm_buf_desc32_t){ + desc.count, desc.size, desc.low_mark, desc.high_mark, + desc.flags, desc.agp_start + }; + if (copy_to_user(argp, &desc32, sizeof(drm_buf_desc32_t))) return -EFAULT; return 0; @@ -408,21 +346,17 @@ static int compat_drm_markbufs(struct file *file, unsigned int cmd, { drm_buf_desc32_t b32; drm_buf_desc32_t __user *argp = (void __user *)arg; - struct drm_buf_desc __user *buf; + struct drm_buf_desc buf; if (copy_from_user(&b32, argp, sizeof(b32))) return -EFAULT; - buf = compat_alloc_user_space(sizeof(*buf)); - if (!buf) - return -EFAULT; - - if (__put_user(b32.size, &buf->size) - || __put_user(b32.low_mark, &buf->low_mark) - || __put_user(b32.high_mark, &buf->high_mark)) - return -EFAULT; + buf.size = b32.size; + buf.low_mark = b32.low_mark; + buf.high_mark = b32.high_mark; - return drm_ioctl(file, DRM_IOCTL_MARK_BUFS, (unsigned long)buf); + return drm_ioctl_kernel(file, drm_legacy_markbufs, &buf, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } typedef struct drm_buf_info32 { @@ -430,52 +364,42 @@ typedef struct drm_buf_info32 { u32 list; } drm_buf_info32_t; +static int copy_one_buf32(void *data, int count, struct drm_buf_entry *from) +{ + drm_buf_info32_t *request = data; + drm_buf_desc32_t __user *to = compat_ptr(request->list); + drm_buf_desc32_t v = {.count = from->buf_count, + .size = from->buf_size, + .low_mark = from->low_mark, + .high_mark = from->high_mark}; + return copy_to_user(to + count, &v, offsetof(drm_buf_desc32_t, flags)); +} + +static int drm_legacy_infobufs32(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + drm_buf_info32_t *request = data; + return __drm_legacy_infobufs(dev, data, &request->count, copy_one_buf32); +} + static int compat_drm_infobufs(struct file *file, unsigned int cmd, unsigned long arg) { drm_buf_info32_t req32; drm_buf_info32_t __user *argp = (void __user *)arg; - drm_buf_desc32_t __user *to; - struct drm_buf_info __user *request; - struct drm_buf_desc __user *list; - size_t nbytes; - int i, err; - int count, actual; + int err; if (copy_from_user(&req32, argp, sizeof(req32))) return -EFAULT; - count = req32.count; - to = (drm_buf_desc32_t __user *) (unsigned long)req32.list; - if (count < 0) - count = 0; - if (count > 0 - && !access_ok(VERIFY_WRITE, to, count * sizeof(drm_buf_desc32_t))) - return -EFAULT; - - nbytes = sizeof(*request) + count * sizeof(struct drm_buf_desc); - request = compat_alloc_user_space(nbytes); - if (!request) - return -EFAULT; - list = (struct drm_buf_desc *) (request + 1); - - if (__put_user(count, &request->count) - || __put_user(list, &request->list)) - return -EFAULT; + if (req32.count < 0) + req32.count = 0; - err = drm_ioctl(file, DRM_IOCTL_INFO_BUFS, (unsigned long)request); + err = drm_ioctl_kernel(file, drm_legacy_infobufs32, &req32, DRM_AUTH); if (err) return err; - if (__get_user(actual, &request->count)) - return -EFAULT; - if (count >= actual) - for (i = 0; i < actual; ++i) - if (__copy_in_user(&to[i], &list[i], - offsetof(struct drm_buf_desc, flags))) - return -EFAULT; - - if (__put_user(actual, &argp->count)) + if (put_user(req32.count, &argp->count)) return -EFAULT; return 0; @@ -494,54 +418,52 @@ typedef struct drm_buf_map32 { u32 list; /**< Buffer information */ } drm_buf_map32_t; +static int map_one_buf32(void *data, int idx, unsigned long virtual, + struct drm_buf *buf) +{ + drm_buf_map32_t *request = data; + drm_buf_pub32_t __user *to = compat_ptr(request->list) + idx; + drm_buf_pub32_t v; + + v.idx = buf->idx; + v.total = buf->total; + v.used = 0; + v.address = virtual + buf->offset; + if (copy_to_user(to, &v, sizeof(v))) + return -EFAULT; + return 0; +} + +static int drm_legacy_mapbufs32(struct drm_device *dev, void *data, + struct drm_file *file_priv) +{ + drm_buf_map32_t *request = data; + void __user *v; + int err = __drm_legacy_mapbufs(dev, data, &request->count, + &v, map_one_buf32, + file_priv); + request->virtual = ptr_to_compat(v); + return err; +} + static int compat_drm_mapbufs(struct file *file, unsigned int cmd, unsigned long arg) { drm_buf_map32_t __user *argp = (void __user *)arg; drm_buf_map32_t req32; - drm_buf_pub32_t __user *list32; - struct drm_buf_map __user *request; - struct drm_buf_pub __user *list; - int i, err; - int count, actual; - size_t nbytes; - void __user *addr; + int err; if (copy_from_user(&req32, argp, sizeof(req32))) return -EFAULT; - count = req32.count; - list32 = (void __user *)(unsigned long)req32.list; - - if (count < 0) + if (req32.count < 0) return -EINVAL; - nbytes = sizeof(*request) + count * sizeof(struct drm_buf_pub); - request = compat_alloc_user_space(nbytes); - if (!request) - return -EFAULT; - list = (struct drm_buf_pub *) (request + 1); - - if (__put_user(count, &request->count) - || __put_user(list, &request->list)) - return -EFAULT; - err = drm_ioctl(file, DRM_IOCTL_MAP_BUFS, (unsigned long)request); + err = drm_ioctl_kernel(file, drm_legacy_mapbufs32, &req32, DRM_AUTH); if (err) return err; - if (__get_user(actual, &request->count)) - return -EFAULT; - if (count >= actual) - for (i = 0; i < actual; ++i) - if (__copy_in_user(&list32[i], &list[i], - offsetof(struct drm_buf_pub, address)) - || __get_user(addr, &list[i].address) - || __put_user((unsigned long)addr, - &list32[i].address)) - return -EFAULT; - - if (__put_user(actual, &argp->count) - || __get_user(addr, &request->virtual) - || __put_user((unsigned long)addr, &argp->virtual)) + if (put_user(req32.count, &argp->count) + || put_user(req32.virtual, &argp->virtual)) return -EFAULT; return 0; @@ -556,21 +478,15 @@ static int compat_drm_freebufs(struct file *file, unsigned int cmd, unsigned long arg) { drm_buf_free32_t req32; - struct drm_buf_free __user *request; + struct drm_buf_free request; drm_buf_free32_t __user *argp = (void __user *)arg; if (copy_from_user(&req32, argp, sizeof(req32))) return -EFAULT; - request = compat_alloc_user_space(sizeof(*request)); - if (!request) - return -EFAULT; - if (__put_user(req32.count, &request->count) - || __put_user((int __user *)(unsigned long)req32.list, - &request->list)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_FREE_BUFS, (unsigned long)request); + request.count = req32.count; + request.list = compat_ptr(req32.list); + return drm_ioctl_kernel(file, drm_legacy_freebufs, &request, DRM_AUTH); } typedef struct drm_ctx_priv_map32 { @@ -582,48 +498,36 @@ static int compat_drm_setsareactx(struct file *file, unsigned int cmd, unsigned long arg) { drm_ctx_priv_map32_t req32; - struct drm_ctx_priv_map __user *request; + struct drm_ctx_priv_map request; drm_ctx_priv_map32_t __user *argp = (void __user *)arg; if (copy_from_user(&req32, argp, sizeof(req32))) return -EFAULT; - request = compat_alloc_user_space(sizeof(*request)); - if (!request) - return -EFAULT; - if (__put_user(req32.ctx_id, &request->ctx_id) - || __put_user((void *)(unsigned long)req32.handle, - &request->handle)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_SET_SAREA_CTX, (unsigned long)request); + request.ctx_id = req32.ctx_id; + request.handle = compat_ptr(req32.handle); + return drm_ioctl_kernel(file, drm_legacy_setsareactx, &request, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } static int compat_drm_getsareactx(struct file *file, unsigned int cmd, unsigned long arg) { - struct drm_ctx_priv_map __user *request; + struct drm_ctx_priv_map req; + drm_ctx_priv_map32_t req32; drm_ctx_priv_map32_t __user *argp = (void __user *)arg; int err; - unsigned int ctx_id; - void *handle; - - if (!access_ok(VERIFY_WRITE, argp, sizeof(*argp)) - || __get_user(ctx_id, &argp->ctx_id)) - return -EFAULT; - request = compat_alloc_user_space(sizeof(*request)); - if (!request) - return -EFAULT; - if (__put_user(ctx_id, &request->ctx_id)) + if (copy_from_user(&req32, argp, sizeof(req32))) return -EFAULT; - err = drm_ioctl(file, DRM_IOCTL_GET_SAREA_CTX, (unsigned long)request); + req.ctx_id = req32.ctx_id; + err = drm_ioctl_kernel(file, drm_legacy_getsareactx, &req, DRM_AUTH); if (err) return err; - if (__get_user(handle, &request->handle) - || __put_user((unsigned long)handle, &argp->handle)) + req32.handle = ptr_to_compat(req.handle); + if (copy_to_user(argp, &req32, sizeof(req32))) return -EFAULT; return 0; @@ -639,26 +543,20 @@ static int compat_drm_resctx(struct file *file, unsigned int cmd, { drm_ctx_res32_t __user *argp = (void __user *)arg; drm_ctx_res32_t res32; - struct drm_ctx_res __user *res; + struct drm_ctx_res res; int err; if (copy_from_user(&res32, argp, sizeof(res32))) return -EFAULT; - res = compat_alloc_user_space(sizeof(*res)); - if (!res) - return -EFAULT; - if (__put_user(res32.count, &res->count) - || __put_user((struct drm_ctx __user *) (unsigned long)res32.contexts, - &res->contexts)) - return -EFAULT; - - err = drm_ioctl(file, DRM_IOCTL_RES_CTX, (unsigned long)res); + res.count = res32.count; + res.contexts = compat_ptr(res32.contexts); + err = drm_ioctl_kernel(file, drm_legacy_resctx, &res, DRM_AUTH); if (err) return err; - if (__get_user(res32.count, &res->count) - || __put_user(res32.count, &argp->count)) + res32.count = res.count; + if (copy_to_user(argp, &res32, sizeof(res32))) return -EFAULT; return 0; @@ -682,38 +580,26 @@ static int compat_drm_dma(struct file *file, unsigned int cmd, { drm_dma32_t d32; drm_dma32_t __user *argp = (void __user *)arg; - struct drm_dma __user *d; + struct drm_dma d; int err; if (copy_from_user(&d32, argp, sizeof(d32))) return -EFAULT; - d = compat_alloc_user_space(sizeof(*d)); - if (!d) - return -EFAULT; - - if (__put_user(d32.context, &d->context) - || __put_user(d32.send_count, &d->send_count) - || __put_user((int __user *)(unsigned long)d32.send_indices, - &d->send_indices) - || __put_user((int __user *)(unsigned long)d32.send_sizes, - &d->send_sizes) - || __put_user(d32.flags, &d->flags) - || __put_user(d32.request_count, &d->request_count) - || __put_user((int __user *)(unsigned long)d32.request_indices, - &d->request_indices) - || __put_user((int __user *)(unsigned long)d32.request_sizes, - &d->request_sizes)) - return -EFAULT; - - err = drm_ioctl(file, DRM_IOCTL_DMA, (unsigned long)d); + d.context = d32.context; + d.send_count = d32.send_count; + d.send_indices = compat_ptr(d32.send_indices); + d.send_sizes = compat_ptr(d32.send_sizes); + d.flags = d32.flags; + d.request_count = d32.request_count; + d.request_indices = compat_ptr(d32.request_indices); + d.request_sizes = compat_ptr(d32.request_sizes); + err = drm_ioctl_kernel(file, drm_legacy_dma_ioctl, &d, DRM_AUTH); if (err) return err; - if (__get_user(d32.request_size, &d->request_size) - || __get_user(d32.granted_count, &d->granted_count) - || __put_user(d32.request_size, &argp->request_size) - || __put_user(d32.granted_count, &argp->granted_count)) + if (put_user(d.request_size, &argp->request_size) + || put_user(d.granted_count, &argp->granted_count)) return -EFAULT; return 0; @@ -728,17 +614,13 @@ static int compat_drm_agp_enable(struct file *file, unsigned int cmd, unsigned long arg) { drm_agp_mode32_t __user *argp = (void __user *)arg; - drm_agp_mode32_t m32; - struct drm_agp_mode __user *mode; - - if (get_user(m32.mode, &argp->mode)) - return -EFAULT; + struct drm_agp_mode mode; - mode = compat_alloc_user_space(sizeof(*mode)); - if (put_user(m32.mode, &mode->mode)) + if (get_user(mode.mode, &argp->mode)) return -EFAULT; - return drm_ioctl(file, DRM_IOCTL_AGP_ENABLE, (unsigned long)mode); + return drm_ioctl_kernel(file, drm_agp_enable_ioctl, &mode, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } typedef struct drm_agp_info32 { @@ -760,28 +642,22 @@ static int compat_drm_agp_info(struct file *file, unsigned int cmd, { drm_agp_info32_t __user *argp = (void __user *)arg; drm_agp_info32_t i32; - struct drm_agp_info __user *info; + struct drm_agp_info info; int err; - info = compat_alloc_user_space(sizeof(*info)); - if (!info) - return -EFAULT; - - err = drm_ioctl(file, DRM_IOCTL_AGP_INFO, (unsigned long)info); + err = drm_ioctl_kernel(file, drm_agp_info_ioctl, &info, DRM_AUTH); if (err) return err; - if (__get_user(i32.agp_version_major, &info->agp_version_major) - || __get_user(i32.agp_version_minor, &info->agp_version_minor) - || __get_user(i32.mode, &info->mode) - || __get_user(i32.aperture_base, &info->aperture_base) - || __get_user(i32.aperture_size, &info->aperture_size) - || __get_user(i32.memory_allowed, &info->memory_allowed) - || __get_user(i32.memory_used, &info->memory_used) - || __get_user(i32.id_vendor, &info->id_vendor) - || __get_user(i32.id_device, &info->id_device)) - return -EFAULT; - + i32.agp_version_major = info.agp_version_major; + i32.agp_version_minor = info.agp_version_minor; + i32.mode = info.mode; + i32.aperture_base = info.aperture_base; + i32.aperture_size = info.aperture_size; + i32.memory_allowed = info.memory_allowed; + i32.memory_used = info.memory_used; + i32.id_vendor = info.id_vendor; + i32.id_device = info.id_device; if (copy_to_user(argp, &i32, sizeof(i32))) return -EFAULT; @@ -800,26 +676,24 @@ static int compat_drm_agp_alloc(struct file *file, unsigned int cmd, { drm_agp_buffer32_t __user *argp = (void __user *)arg; drm_agp_buffer32_t req32; - struct drm_agp_buffer __user *request; + struct drm_agp_buffer request; int err; if (copy_from_user(&req32, argp, sizeof(req32))) return -EFAULT; - request = compat_alloc_user_space(sizeof(*request)); - if (!request - || __put_user(req32.size, &request->size) - || __put_user(req32.type, &request->type)) - return -EFAULT; - - err = drm_ioctl(file, DRM_IOCTL_AGP_ALLOC, (unsigned long)request); + request.size = req32.size; + request.type = req32.type; + err = drm_ioctl_kernel(file, drm_agp_alloc_ioctl, &request, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); if (err) return err; - if (__get_user(req32.handle, &request->handle) - || __get_user(req32.physical, &request->physical) - || copy_to_user(argp, &req32, sizeof(req32))) { - drm_ioctl(file, DRM_IOCTL_AGP_FREE, (unsigned long)request); + req32.handle = request.handle; + req32.physical = request.physical; + if (copy_to_user(argp, &req32, sizeof(req32))) { + drm_ioctl_kernel(file, drm_agp_free_ioctl, &request, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); return -EFAULT; } @@ -830,16 +704,13 @@ static int compat_drm_agp_free(struct file *file, unsigned int cmd, unsigned long arg) { drm_agp_buffer32_t __user *argp = (void __user *)arg; - struct drm_agp_buffer __user *request; - u32 handle; + struct drm_agp_buffer request; - request = compat_alloc_user_space(sizeof(*request)); - if (!request - || get_user(handle, &argp->handle) - || __put_user(handle, &request->handle)) + if (get_user(request.handle, &argp->handle)) return -EFAULT; - return drm_ioctl(file, DRM_IOCTL_AGP_FREE, (unsigned long)request); + return drm_ioctl_kernel(file, drm_agp_free_ioctl, &request, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } typedef struct drm_agp_binding32 { @@ -852,34 +723,28 @@ static int compat_drm_agp_bind(struct file *file, unsigned int cmd, { drm_agp_binding32_t __user *argp = (void __user *)arg; drm_agp_binding32_t req32; - struct drm_agp_binding __user *request; + struct drm_agp_binding request; if (copy_from_user(&req32, argp, sizeof(req32))) return -EFAULT; - request = compat_alloc_user_space(sizeof(*request)); - if (!request - || __put_user(req32.handle, &request->handle) - || __put_user(req32.offset, &request->offset)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_AGP_BIND, (unsigned long)request); + request.handle = req32.handle; + request.offset = req32.offset; + return drm_ioctl_kernel(file, drm_agp_bind_ioctl, &request, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } static int compat_drm_agp_unbind(struct file *file, unsigned int cmd, unsigned long arg) { drm_agp_binding32_t __user *argp = (void __user *)arg; - struct drm_agp_binding __user *request; - u32 handle; + struct drm_agp_binding request; - request = compat_alloc_user_space(sizeof(*request)); - if (!request - || get_user(handle, &argp->handle) - || __put_user(handle, &request->handle)) + if (get_user(request.handle, &argp->handle)) return -EFAULT; - return drm_ioctl(file, DRM_IOCTL_AGP_UNBIND, (unsigned long)request); + return drm_ioctl_kernel(file, drm_agp_unbind_ioctl, &request, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } #endif /* CONFIG_AGP */ @@ -892,23 +757,19 @@ static int compat_drm_sg_alloc(struct file *file, unsigned int cmd, unsigned long arg) { drm_scatter_gather32_t __user *argp = (void __user *)arg; - struct drm_scatter_gather __user *request; + struct drm_scatter_gather request; int err; - unsigned long x; - request = compat_alloc_user_space(sizeof(*request)); - if (!request || !access_ok(VERIFY_WRITE, argp, sizeof(*argp)) - || __get_user(x, &argp->size) - || __put_user(x, &request->size)) + if (get_user(request.size, &argp->size)) return -EFAULT; - err = drm_ioctl(file, DRM_IOCTL_SG_ALLOC, (unsigned long)request); + err = drm_ioctl_kernel(file, drm_legacy_sg_alloc, &request, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); if (err) return err; /* XXX not sure about the handle conversion here... */ - if (__get_user(x, &request->handle) - || __put_user(x >> PAGE_SHIFT, &argp->handle)) + if (put_user(request.handle >> PAGE_SHIFT, &argp->handle)) return -EFAULT; return 0; @@ -918,19 +779,17 @@ static int compat_drm_sg_free(struct file *file, unsigned int cmd, unsigned long arg) { drm_scatter_gather32_t __user *argp = (void __user *)arg; - struct drm_scatter_gather __user *request; + struct drm_scatter_gather request; unsigned long x; - request = compat_alloc_user_space(sizeof(*request)); - if (!request || !access_ok(VERIFY_WRITE, argp, sizeof(*argp)) - || __get_user(x, &argp->handle) - || __put_user(x << PAGE_SHIFT, &request->handle)) + if (get_user(x, &argp->handle)) return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_SG_FREE, (unsigned long)request); + request.handle = x << PAGE_SHIFT; + return drm_ioctl_kernel(file, drm_legacy_sg_free, &request, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } -#if defined(CONFIG_X86) || defined(CONFIG_IA64) +#if defined(CONFIG_X86) typedef struct drm_update_draw32 { drm_drawable_t handle; unsigned int type; @@ -943,22 +802,11 @@ static int compat_drm_update_draw(struct file *file, unsigned int cmd, unsigned long arg) { drm_update_draw32_t update32; - struct drm_update_draw __user *request; - int err; - if (copy_from_user(&update32, (void __user *)arg, sizeof(update32))) return -EFAULT; - request = compat_alloc_user_space(sizeof(*request)); - if (!request || - __put_user(update32.handle, &request->handle) || - __put_user(update32.type, &request->type) || - __put_user(update32.num, &request->num) || - __put_user(update32.data, &request->data)) - return -EFAULT; - - err = drm_ioctl(file, DRM_IOCTL_UPDATE_DRAW, (unsigned long)request); - return err; + return drm_ioctl_kernel(file, drm_noop, NULL, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } #endif @@ -985,36 +833,30 @@ static int compat_drm_wait_vblank(struct file *file, unsigned int cmd, { drm_wait_vblank32_t __user *argp = (void __user *)arg; drm_wait_vblank32_t req32; - union drm_wait_vblank __user *request; + union drm_wait_vblank req; int err; if (copy_from_user(&req32, argp, sizeof(req32))) return -EFAULT; - request = compat_alloc_user_space(sizeof(*request)); - if (!request - || __put_user(req32.request.type, &request->request.type) - || __put_user(req32.request.sequence, &request->request.sequence) - || __put_user(req32.request.signal, &request->request.signal)) - return -EFAULT; - - err = drm_ioctl(file, DRM_IOCTL_WAIT_VBLANK, (unsigned long)request); + req.request.type = req32.request.type; + req.request.sequence = req32.request.sequence; + req.request.signal = req32.request.signal; + err = drm_ioctl_kernel(file, drm_wait_vblank_ioctl, &req, DRM_UNLOCKED); if (err) return err; - if (__get_user(req32.reply.type, &request->reply.type) - || __get_user(req32.reply.sequence, &request->reply.sequence) - || __get_user(req32.reply.tval_sec, &request->reply.tval_sec) - || __get_user(req32.reply.tval_usec, &request->reply.tval_usec)) - return -EFAULT; - + req32.reply.type = req.reply.type; + req32.reply.sequence = req.reply.sequence; + req32.reply.tval_sec = req.reply.tval_sec; + req32.reply.tval_usec = req.reply.tval_usec; if (copy_to_user(argp, &req32, sizeof(req32))) return -EFAULT; return 0; } -#if defined(CONFIG_X86) || defined(CONFIG_IA64) +#if defined(CONFIG_X86) typedef struct drm_mode_fb_cmd232 { u32 fb_id; u32 width; @@ -1031,82 +873,67 @@ static int compat_drm_mode_addfb2(struct file *file, unsigned int cmd, unsigned long arg) { struct drm_mode_fb_cmd232 __user *argp = (void __user *)arg; - struct drm_mode_fb_cmd232 req32; - struct drm_mode_fb_cmd2 __user *req64; - int i; + struct drm_mode_fb_cmd2 req64; int err; - if (copy_from_user(&req32, argp, sizeof(req32))) + if (copy_from_user(&req64, argp, + offsetof(drm_mode_fb_cmd232_t, modifier))) return -EFAULT; - req64 = compat_alloc_user_space(sizeof(*req64)); - - if (!access_ok(VERIFY_WRITE, req64, sizeof(*req64)) - || __put_user(req32.width, &req64->width) - || __put_user(req32.height, &req64->height) - || __put_user(req32.pixel_format, &req64->pixel_format) - || __put_user(req32.flags, &req64->flags)) + if (copy_from_user(&req64.modifier, &argp->modifier, + sizeof(req64.modifier))) return -EFAULT; - for (i = 0; i < 4; i++) { - if (__put_user(req32.handles[i], &req64->handles[i])) - return -EFAULT; - if (__put_user(req32.pitches[i], &req64->pitches[i])) - return -EFAULT; - if (__put_user(req32.offsets[i], &req64->offsets[i])) - return -EFAULT; - if (__put_user(req32.modifier[i], &req64->modifier[i])) - return -EFAULT; - } - - err = drm_ioctl(file, DRM_IOCTL_MODE_ADDFB2, (unsigned long)req64); + err = drm_ioctl_kernel(file, drm_mode_addfb2, &req64, + DRM_CONTROL_ALLOW|DRM_UNLOCKED); if (err) return err; - if (__get_user(req32.fb_id, &req64->fb_id)) - return -EFAULT; - - if (copy_to_user(argp, &req32, sizeof(req32))) + if (put_user(req64.fb_id, &argp->fb_id)) return -EFAULT; return 0; } #endif -static drm_ioctl_compat_t *drm_compat_ioctls[] = { - [DRM_IOCTL_NR(DRM_IOCTL_VERSION32)] = compat_drm_version, - [DRM_IOCTL_NR(DRM_IOCTL_GET_UNIQUE32)] = compat_drm_getunique, - [DRM_IOCTL_NR(DRM_IOCTL_GET_MAP32)] = compat_drm_getmap, - [DRM_IOCTL_NR(DRM_IOCTL_GET_CLIENT32)] = compat_drm_getclient, - [DRM_IOCTL_NR(DRM_IOCTL_GET_STATS32)] = compat_drm_getstats, - [DRM_IOCTL_NR(DRM_IOCTL_SET_UNIQUE32)] = compat_drm_setunique, - [DRM_IOCTL_NR(DRM_IOCTL_ADD_MAP32)] = compat_drm_addmap, - [DRM_IOCTL_NR(DRM_IOCTL_ADD_BUFS32)] = compat_drm_addbufs, - [DRM_IOCTL_NR(DRM_IOCTL_MARK_BUFS32)] = compat_drm_markbufs, - [DRM_IOCTL_NR(DRM_IOCTL_INFO_BUFS32)] = compat_drm_infobufs, - [DRM_IOCTL_NR(DRM_IOCTL_MAP_BUFS32)] = compat_drm_mapbufs, - [DRM_IOCTL_NR(DRM_IOCTL_FREE_BUFS32)] = compat_drm_freebufs, - [DRM_IOCTL_NR(DRM_IOCTL_RM_MAP32)] = compat_drm_rmmap, - [DRM_IOCTL_NR(DRM_IOCTL_SET_SAREA_CTX32)] = compat_drm_setsareactx, - [DRM_IOCTL_NR(DRM_IOCTL_GET_SAREA_CTX32)] = compat_drm_getsareactx, - [DRM_IOCTL_NR(DRM_IOCTL_RES_CTX32)] = compat_drm_resctx, - [DRM_IOCTL_NR(DRM_IOCTL_DMA32)] = compat_drm_dma, +static struct { + drm_ioctl_compat_t *fn; + char *name; +} drm_compat_ioctls[] = { +#define DRM_IOCTL32_DEF(n, f) [DRM_IOCTL_NR(n##32)] = {.fn = f, .name = #n} + DRM_IOCTL32_DEF(DRM_IOCTL_VERSION, compat_drm_version), + DRM_IOCTL32_DEF(DRM_IOCTL_GET_UNIQUE, compat_drm_getunique), + DRM_IOCTL32_DEF(DRM_IOCTL_GET_MAP, compat_drm_getmap), + DRM_IOCTL32_DEF(DRM_IOCTL_GET_CLIENT, compat_drm_getclient), + DRM_IOCTL32_DEF(DRM_IOCTL_GET_STATS, compat_drm_getstats), + DRM_IOCTL32_DEF(DRM_IOCTL_SET_UNIQUE, compat_drm_setunique), + DRM_IOCTL32_DEF(DRM_IOCTL_ADD_MAP, compat_drm_addmap), + DRM_IOCTL32_DEF(DRM_IOCTL_ADD_BUFS, compat_drm_addbufs), + DRM_IOCTL32_DEF(DRM_IOCTL_MARK_BUFS, compat_drm_markbufs), + DRM_IOCTL32_DEF(DRM_IOCTL_INFO_BUFS, compat_drm_infobufs), + DRM_IOCTL32_DEF(DRM_IOCTL_MAP_BUFS, compat_drm_mapbufs), + DRM_IOCTL32_DEF(DRM_IOCTL_FREE_BUFS, compat_drm_freebufs), + DRM_IOCTL32_DEF(DRM_IOCTL_RM_MAP, compat_drm_rmmap), + DRM_IOCTL32_DEF(DRM_IOCTL_SET_SAREA_CTX, compat_drm_setsareactx), + DRM_IOCTL32_DEF(DRM_IOCTL_GET_SAREA_CTX, compat_drm_getsareactx), + DRM_IOCTL32_DEF(DRM_IOCTL_RES_CTX, compat_drm_resctx), + DRM_IOCTL32_DEF(DRM_IOCTL_DMA, compat_drm_dma), #if IS_ENABLED(CONFIG_AGP) - [DRM_IOCTL_NR(DRM_IOCTL_AGP_ENABLE32)] = compat_drm_agp_enable, - [DRM_IOCTL_NR(DRM_IOCTL_AGP_INFO32)] = compat_drm_agp_info, - [DRM_IOCTL_NR(DRM_IOCTL_AGP_ALLOC32)] = compat_drm_agp_alloc, - [DRM_IOCTL_NR(DRM_IOCTL_AGP_FREE32)] = compat_drm_agp_free, - [DRM_IOCTL_NR(DRM_IOCTL_AGP_BIND32)] = compat_drm_agp_bind, - [DRM_IOCTL_NR(DRM_IOCTL_AGP_UNBIND32)] = compat_drm_agp_unbind, + DRM_IOCTL32_DEF(DRM_IOCTL_AGP_ENABLE, compat_drm_agp_enable), + DRM_IOCTL32_DEF(DRM_IOCTL_AGP_INFO, compat_drm_agp_info), + DRM_IOCTL32_DEF(DRM_IOCTL_AGP_ALLOC, compat_drm_agp_alloc), + DRM_IOCTL32_DEF(DRM_IOCTL_AGP_FREE, compat_drm_agp_free), + DRM_IOCTL32_DEF(DRM_IOCTL_AGP_BIND, compat_drm_agp_bind), + DRM_IOCTL32_DEF(DRM_IOCTL_AGP_UNBIND, compat_drm_agp_unbind), #endif - [DRM_IOCTL_NR(DRM_IOCTL_SG_ALLOC32)] = compat_drm_sg_alloc, - [DRM_IOCTL_NR(DRM_IOCTL_SG_FREE32)] = compat_drm_sg_free, + DRM_IOCTL32_DEF(DRM_IOCTL_SG_ALLOC, compat_drm_sg_alloc), + DRM_IOCTL32_DEF(DRM_IOCTL_SG_FREE, compat_drm_sg_free), #if defined(CONFIG_X86) || defined(CONFIG_IA64) - [DRM_IOCTL_NR(DRM_IOCTL_UPDATE_DRAW32)] = compat_drm_update_draw, + DRM_IOCTL32_DEF(DRM_IOCTL_UPDATE_DRAW, compat_drm_update_draw), #endif - [DRM_IOCTL_NR(DRM_IOCTL_WAIT_VBLANK32)] = compat_drm_wait_vblank, + DRM_IOCTL32_DEF(DRM_IOCTL_WAIT_VBLANK, compat_drm_wait_vblank), #if defined(CONFIG_X86) || defined(CONFIG_IA64) - [DRM_IOCTL_NR(DRM_IOCTL_MODE_ADDFB232)] = compat_drm_mode_addfb2, + DRM_IOCTL32_DEF(DRM_IOCTL_MODE_ADDFB2, compat_drm_mode_addfb2), #endif }; @@ -1127,6 +954,7 @@ static drm_ioctl_compat_t *drm_compat_ioctls[] = { long drm_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { unsigned int nr = DRM_IOCTL_NR(cmd); + struct drm_file *file_priv = filp->private_data; drm_ioctl_compat_t *fn; int ret; @@ -1137,13 +965,18 @@ long drm_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) if (nr >= ARRAY_SIZE(drm_compat_ioctls)) return drm_ioctl(filp, cmd, arg); - fn = drm_compat_ioctls[nr]; - - if (fn != NULL) - ret = (*fn) (filp, cmd, arg); - else - ret = drm_ioctl(filp, cmd, arg); + fn = drm_compat_ioctls[nr].fn; + if (!fn) + return drm_ioctl(filp, cmd, arg); + DRM_DEBUG("pid=%d, dev=0x%lx, auth=%d, %s\n", + task_pid_nr(current), + (long)old_encode_dev(file_priv->minor->kdev->devt), + file_priv->authenticated, + drm_compat_ioctls[nr].name); + ret = (*fn)(filp, cmd, arg); + if (ret) + DRM_DEBUG("ret = %d\n", ret); return ret; } EXPORT_SYMBOL(drm_compat_ioctl); diff --git a/drivers/gpu/drm/drm_ioctl.c b/drivers/gpu/drm/drm_ioctl.c index f7f150e4a0c0..8bfeb32f8a10 100644 --- a/drivers/gpu/drm/drm_ioctl.c +++ b/drivers/gpu/drm/drm_ioctl.c @@ -107,7 +107,7 @@ * * Copies the bus id from drm_device::unique into user space. */ -static int drm_getunique(struct drm_device *dev, void *data, +int drm_getunique(struct drm_device *dev, void *data, struct drm_file *file_priv) { struct drm_unique *u = data; @@ -172,7 +172,7 @@ static int drm_set_busid(struct drm_device *dev, struct drm_file *file_priv) * Searches for the client with the specified index and copies its information * into userspace */ -static int drm_getclient(struct drm_device *dev, void *data, +int drm_getclient(struct drm_device *dev, void *data, struct drm_file *file_priv) { struct drm_client *client = data; @@ -464,7 +464,7 @@ static int drm_copy_field(char __user *buf, size_t *buf_len, const char *value) * * Fills in the version information in \p arg. */ -static int drm_version(struct drm_device *dev, void *data, +int drm_version(struct drm_device *dev, void *data, struct drm_file *file_priv) { struct drm_version *version = data; @@ -709,6 +709,33 @@ static const struct drm_ioctl_desc drm_ioctls[] = { * the driver-specific IOCTLs are wired up. */ +long drm_ioctl_kernel(struct file *file, drm_ioctl_t *func, void *kdata, + u32 flags) +{ + struct drm_file *file_priv = file->private_data; + struct drm_device *dev = file_priv->minor->dev; + int retcode; + + if (drm_device_is_unplugged(dev)) + return -ENODEV; + + retcode = drm_ioctl_permit(flags, file_priv); + if (unlikely(retcode)) + return retcode; + + /* Enforce sane locking for modern driver ioctls. */ + if (!drm_core_check_feature(dev, DRIVER_LEGACY) || + (flags & DRM_UNLOCKED)) + retcode = func(dev, kdata, file_priv); + else { + mutex_lock(&drm_global_mutex); + retcode = func(dev, kdata, file_priv); + mutex_unlock(&drm_global_mutex); + } + return retcode; +} +EXPORT_SYMBOL(drm_ioctl_kernel); + /** * drm_ioctl - ioctl callback implementation for DRM drivers * @filp: file this ioctl is called on @@ -777,10 +804,6 @@ long drm_ioctl(struct file *filp, goto err_i1; } - retcode = drm_ioctl_permit(ioctl->flags, file_priv); - if (unlikely(retcode)) - goto err_i1; - if (ksize <= sizeof(stack_kdata)) { kdata = stack_kdata; } else { @@ -799,16 +822,7 @@ long drm_ioctl(struct file *filp, if (ksize > in_size) memset(kdata + in_size, 0, ksize - in_size); - /* Enforce sane locking for modern driver ioctls. */ - if (!drm_core_check_feature(dev, DRIVER_LEGACY) || - (ioctl->flags & DRM_UNLOCKED)) - retcode = func(dev, kdata, file_priv); - else { - mutex_lock(&drm_global_mutex); - retcode = func(dev, kdata, file_priv); - mutex_unlock(&drm_global_mutex); - } - + retcode = drm_ioctl_kernel(filp, func, kdata, ioctl->flags); if (copy_to_user((void __user *)arg, kdata, out_size) != 0) retcode = -EFAULT; diff --git a/drivers/gpu/drm/drm_legacy.h b/drivers/gpu/drm/drm_legacy.h index e4bb5ad747c8..280fbeb846ff 100644 --- a/drivers/gpu/drm/drm_legacy.h +++ b/drivers/gpu/drm/drm_legacy.h @@ -74,6 +74,13 @@ int drm_legacy_freebufs(struct drm_device *d, void *v, struct drm_file *f); int drm_legacy_mapbufs(struct drm_device *d, void *v, struct drm_file *f); int drm_legacy_dma_ioctl(struct drm_device *d, void *v, struct drm_file *f); +int __drm_legacy_infobufs(struct drm_device *, void *, int *, + int (*)(void *, int, struct drm_buf_entry *)); +int __drm_legacy_mapbufs(struct drm_device *, void *, int *, + void __user **, + int (*)(void *, int, unsigned long, struct drm_buf *), + struct drm_file *); + #ifdef CONFIG_DRM_VM void drm_legacy_vma_flush(struct drm_device *d); #else diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c index f64af5e06ac6..a5b38a80a99a 100644 --- a/drivers/gpu/drm/drm_syncobj.c +++ b/drivers/gpu/drm/drm_syncobj.c @@ -77,17 +77,15 @@ EXPORT_SYMBOL(drm_syncobj_find); /** * drm_syncobj_replace_fence - replace fence in a sync object. - * @file_private: drm file private pointer. * @syncobj: Sync object to replace fence in * @fence: fence to install in sync file. * * This replaces the fence on a sync object. */ -void drm_syncobj_replace_fence(struct drm_file *file_private, - struct drm_syncobj *syncobj, +void drm_syncobj_replace_fence(struct drm_syncobj *syncobj, struct dma_fence *fence) { - struct dma_fence *old_fence = NULL; + struct dma_fence *old_fence; if (fence) dma_fence_get(fence); @@ -292,7 +290,7 @@ int drm_syncobj_import_sync_file_fence(struct drm_file *file_private, return -ENOENT; } - drm_syncobj_replace_fence(file_private, syncobj, fence); + drm_syncobj_replace_fence(syncobj, fence); dma_fence_put(fence); drm_syncobj_put(syncobj); return 0; diff --git a/drivers/gpu/drm/drm_vblank.c b/drivers/gpu/drm/drm_vblank.c index 8099574c8a11..70f2b9593edc 100644 --- a/drivers/gpu/drm/drm_vblank.c +++ b/drivers/gpu/drm/drm_vblank.c @@ -277,7 +277,7 @@ static void drm_update_vblank_count(struct drm_device *dev, unsigned int pipe, * Otherwise reinitialize delayed at next vblank interrupt and assign 0 * for now, to mark the vblanktimestamp as invalid. */ - if (!rc && in_vblank_irq) + if (!rc && !in_vblank_irq) t_vblank = (struct timeval) {0, 0}; store_vblank(dev, pipe, diff, &t_vblank, cur_vblank); diff --git a/drivers/gpu/drm/etnaviv/common.xml.h b/drivers/gpu/drm/etnaviv/common.xml.h index e881482b5971..207f45c999c3 100644 --- a/drivers/gpu/drm/etnaviv/common.xml.h +++ b/drivers/gpu/drm/etnaviv/common.xml.h @@ -8,10 +8,38 @@ http://0x04.net/cgit/index.cgi/rules-ng-ng git clone git://0x04.net/rules-ng-ng The rules-ng-ng source files this header was generated from are: -- state_hi.xml ( 24309 bytes, from 2015-12-12 09:02:53) -- common.xml ( 18379 bytes, from 2015-12-12 09:02:53) +- state.xml ( 19930 bytes, from 2017-03-09 15:43:43) +- common.xml ( 23473 bytes, from 2017-03-09 15:43:43) +- state_hi.xml ( 26403 bytes, from 2017-03-09 15:43:43) +- copyright.xml ( 1597 bytes, from 2016-12-08 16:37:56) +- state_2d.xml ( 51552 bytes, from 2016-12-08 16:37:56) +- state_3d.xml ( 66957 bytes, from 2017-03-09 15:43:43) +- state_vg.xml ( 5975 bytes, from 2016-12-08 16:37:56) -Copyright (C) 2015 +Copyright (C) 2012-2017 by the following authors: +- Wladimir J. van der Laan <laanwj@gmail.com> +- Christian Gmeiner <christian.gmeiner@gmail.com> +- Lucas Stach <l.stach@pengutronix.de> +- Russell King <rmk@arm.linux.org.uk> + +Permission is hereby granted, free of charge, to any person obtaining a +copy of this software and associated documentation files (the "Software"), +to deal in the Software without restriction, including without limitation +the rights to use, copy, modify, merge, publish, distribute, sub license, +and/or sell copies of the Software, and to permit persons to whom the +Software is furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice (including the +next paragraph) shall be included in all copies or substantial portions +of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL +THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING +FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER +DEALINGS IN THE SOFTWARE. */ @@ -162,129 +190,129 @@ Copyright (C) 2015 #define chipMinorFeatures1_FC_FLUSH_STALL 0x80000000 #define chipMinorFeatures2_LINE_LOOP 0x00000001 #define chipMinorFeatures2_LOGIC_OP 0x00000002 -#define chipMinorFeatures2_UNK2 0x00000004 +#define chipMinorFeatures2_SEAMLESS_CUBE_MAP 0x00000004 #define chipMinorFeatures2_SUPERTILED_TEXTURE 0x00000008 -#define chipMinorFeatures2_UNK4 0x00000010 +#define chipMinorFeatures2_LINEAR_PE 0x00000010 #define chipMinorFeatures2_RECT_PRIMITIVE 0x00000020 #define chipMinorFeatures2_COMPOSITION 0x00000040 #define chipMinorFeatures2_CORRECT_AUTO_DISABLE_COUNT 0x00000080 -#define chipMinorFeatures2_UNK8 0x00000100 -#define chipMinorFeatures2_UNK9 0x00000200 -#define chipMinorFeatures2_UNK10 0x00000400 +#define chipMinorFeatures2_PE_SWIZZLE 0x00000100 +#define chipMinorFeatures2_END_EVENT 0x00000200 +#define chipMinorFeatures2_S1S8 0x00000400 #define chipMinorFeatures2_HALTI1 0x00000800 -#define chipMinorFeatures2_UNK12 0x00001000 -#define chipMinorFeatures2_UNK13 0x00002000 -#define chipMinorFeatures2_UNK14 0x00004000 +#define chipMinorFeatures2_RGB888 0x00001000 +#define chipMinorFeatures2_TX__YUV_ASSEMBLER 0x00002000 +#define chipMinorFeatures2_DYNAMIC_FREQUENCY_SCALING 0x00004000 #define chipMinorFeatures2_EXTRA_TEXTURE_STATE 0x00008000 #define chipMinorFeatures2_FULL_DIRECTFB 0x00010000 #define chipMinorFeatures2_2D_TILING 0x00020000 #define chipMinorFeatures2_THREAD_WALKER_IN_PS 0x00040000 #define chipMinorFeatures2_TILE_FILLER 0x00080000 -#define chipMinorFeatures2_UNK20 0x00100000 +#define chipMinorFeatures2_YUV_STANDARD 0x00100000 #define chipMinorFeatures2_2D_MULTI_SOURCE_BLIT 0x00200000 -#define chipMinorFeatures2_UNK22 0x00400000 -#define chipMinorFeatures2_UNK23 0x00800000 -#define chipMinorFeatures2_UNK24 0x01000000 +#define chipMinorFeatures2_YUV_CONVERSION 0x00400000 +#define chipMinorFeatures2_FLUSH_FIXED_2D 0x00800000 +#define chipMinorFeatures2_INTERLEAVER 0x01000000 #define chipMinorFeatures2_MIXED_STREAMS 0x02000000 #define chipMinorFeatures2_2D_420_L2CACHE 0x04000000 -#define chipMinorFeatures2_UNK27 0x08000000 +#define chipMinorFeatures2_BUG_FIXES7 0x08000000 #define chipMinorFeatures2_2D_NO_INDEX8_BRUSH 0x10000000 #define chipMinorFeatures2_TEXTURE_TILED_READ 0x20000000 -#define chipMinorFeatures2_UNK30 0x40000000 -#define chipMinorFeatures2_UNK31 0x80000000 +#define chipMinorFeatures2_DECOMPRESS_Z16 0x40000000 +#define chipMinorFeatures2_BUG_FIXES8 0x80000000 #define chipMinorFeatures3_ROTATION_STALL_FIX 0x00000001 -#define chipMinorFeatures3_UNK1 0x00000002 +#define chipMinorFeatures3_OCL_ONLY 0x00000002 #define chipMinorFeatures3_2D_MULTI_SOURCE_BLT_EX 0x00000004 -#define chipMinorFeatures3_UNK3 0x00000008 -#define chipMinorFeatures3_UNK4 0x00000010 -#define chipMinorFeatures3_UNK5 0x00000020 -#define chipMinorFeatures3_UNK6 0x00000040 -#define chipMinorFeatures3_UNK7 0x00000080 +#define chipMinorFeatures3_INSTRUCTION_CACHE 0x00000008 +#define chipMinorFeatures3_GEOMETRY_SHADER 0x00000010 +#define chipMinorFeatures3_TEX_COMPRESSION_SUPERTILED 0x00000020 +#define chipMinorFeatures3_GENERICS 0x00000040 +#define chipMinorFeatures3_BUG_FIXES9 0x00000080 #define chipMinorFeatures3_FAST_MSAA 0x00000100 -#define chipMinorFeatures3_UNK9 0x00000200 +#define chipMinorFeatures3_WCLIP 0x00000200 #define chipMinorFeatures3_BUG_FIXES10 0x00000400 -#define chipMinorFeatures3_UNK11 0x00000800 +#define chipMinorFeatures3_UNIFIED_SAMPLERS 0x00000800 #define chipMinorFeatures3_BUG_FIXES11 0x00001000 -#define chipMinorFeatures3_UNK13 0x00002000 -#define chipMinorFeatures3_UNK14 0x00004000 -#define chipMinorFeatures3_UNK15 0x00008000 -#define chipMinorFeatures3_UNK16 0x00010000 -#define chipMinorFeatures3_UNK17 0x00020000 +#define chipMinorFeatures3_PERFORMANCE_COUNTERS 0x00002000 +#define chipMinorFeatures3_HAS_FAST_TRANSCENDENTALS 0x00004000 +#define chipMinorFeatures3_BUG_FIXES12 0x00008000 +#define chipMinorFeatures3_BUG_FIXES13 0x00010000 +#define chipMinorFeatures3_DE_ENHANCEMENTS1 0x00020000 #define chipMinorFeatures3_ACE 0x00040000 -#define chipMinorFeatures3_UNK19 0x00080000 -#define chipMinorFeatures3_UNK20 0x00100000 -#define chipMinorFeatures3_UNK21 0x00200000 +#define chipMinorFeatures3_TX_ENHANCEMENTS1 0x00080000 +#define chipMinorFeatures3_SH_ENHANCEMENTS1 0x00100000 +#define chipMinorFeatures3_SH_ENHANCEMENTS2 0x00200000 #define chipMinorFeatures3_UNK22 0x00400000 -#define chipMinorFeatures3_UNK23 0x00800000 +#define chipMinorFeatures3_2D_FC_SOURCE 0x00800000 #define chipMinorFeatures3_UNK24 0x01000000 #define chipMinorFeatures3_UNK25 0x02000000 #define chipMinorFeatures3_NEW_HZ 0x04000000 #define chipMinorFeatures3_UNK27 0x08000000 #define chipMinorFeatures3_UNK28 0x10000000 -#define chipMinorFeatures3_UNK29 0x20000000 +#define chipMinorFeatures3_SH_ENHANCEMENTS3 0x20000000 #define chipMinorFeatures3_UNK30 0x40000000 #define chipMinorFeatures3_UNK31 0x80000000 #define chipMinorFeatures4_UNK0 0x00000001 -#define chipMinorFeatures4_UNK1 0x00000002 -#define chipMinorFeatures4_UNK2 0x00000004 +#define chipMinorFeatures4_PE_ENHANCEMENTS2 0x00000002 +#define chipMinorFeatures4_FRUSTUM_CLIP_FIX 0x00000004 #define chipMinorFeatures4_UNK3 0x00000008 #define chipMinorFeatures4_UNK4 0x00000010 -#define chipMinorFeatures4_UNK5 0x00000020 -#define chipMinorFeatures4_UNK6 0x00000040 +#define chipMinorFeatures4_2D_GAMMA 0x00000020 +#define chipMinorFeatures4_SINGLE_BUFFER 0x00000040 #define chipMinorFeatures4_UNK7 0x00000080 #define chipMinorFeatures4_UNK8 0x00000100 #define chipMinorFeatures4_UNK9 0x00000200 #define chipMinorFeatures4_UNK10 0x00000400 -#define chipMinorFeatures4_UNK11 0x00000800 -#define chipMinorFeatures4_UNK12 0x00001000 -#define chipMinorFeatures4_UNK13 0x00002000 +#define chipMinorFeatures4_TX_LERP_PRECISION_FIX 0x00000800 +#define chipMinorFeatures4_2D_COLOR_SPACE_CONVERSION 0x00001000 +#define chipMinorFeatures4_TEXTURE_ASTC 0x00002000 #define chipMinorFeatures4_UNK14 0x00004000 #define chipMinorFeatures4_UNK15 0x00008000 #define chipMinorFeatures4_HALTI2 0x00010000 #define chipMinorFeatures4_UNK17 0x00020000 #define chipMinorFeatures4_SMALL_MSAA 0x00040000 #define chipMinorFeatures4_UNK19 0x00080000 -#define chipMinorFeatures4_UNK20 0x00100000 -#define chipMinorFeatures4_UNK21 0x00200000 -#define chipMinorFeatures4_UNK22 0x00400000 -#define chipMinorFeatures4_UNK23 0x00800000 -#define chipMinorFeatures4_UNK24 0x01000000 -#define chipMinorFeatures4_UNK25 0x02000000 -#define chipMinorFeatures4_UNK26 0x04000000 -#define chipMinorFeatures4_UNK27 0x08000000 +#define chipMinorFeatures4_NEW_RA 0x00100000 +#define chipMinorFeatures4_2D_OPF_YUV_OUTPUT 0x00200000 +#define chipMinorFeatures4_2D_MULTI_SOURCE_BLT_EX2 0x00400000 +#define chipMinorFeatures4_NO_USER_CSC 0x00800000 +#define chipMinorFeatures4_ZFIXES 0x01000000 +#define chipMinorFeatures4_BUG_FIXES18 0x02000000 +#define chipMinorFeatures4_2D_COMPRESSION 0x04000000 +#define chipMinorFeatures4_PROBE 0x08000000 #define chipMinorFeatures4_UNK28 0x10000000 -#define chipMinorFeatures4_UNK29 0x20000000 +#define chipMinorFeatures4_2D_SUPER_TILE_VERSION 0x20000000 #define chipMinorFeatures4_UNK30 0x40000000 #define chipMinorFeatures4_UNK31 0x80000000 #define chipMinorFeatures5_UNK0 0x00000001 #define chipMinorFeatures5_UNK1 0x00000002 #define chipMinorFeatures5_UNK2 0x00000004 #define chipMinorFeatures5_UNK3 0x00000008 -#define chipMinorFeatures5_UNK4 0x00000010 +#define chipMinorFeatures5_EEZ 0x00000010 #define chipMinorFeatures5_UNK5 0x00000020 #define chipMinorFeatures5_UNK6 0x00000040 #define chipMinorFeatures5_UNK7 0x00000080 #define chipMinorFeatures5_UNK8 0x00000100 #define chipMinorFeatures5_HALTI3 0x00000200 #define chipMinorFeatures5_UNK10 0x00000400 -#define chipMinorFeatures5_UNK11 0x00000800 +#define chipMinorFeatures5_2D_ONE_PASS_FILTER_TAP 0x00000800 #define chipMinorFeatures5_UNK12 0x00001000 -#define chipMinorFeatures5_UNK13 0x00002000 -#define chipMinorFeatures5_UNK14 0x00004000 +#define chipMinorFeatures5_SEPARATE_SRC_DST 0x00002000 +#define chipMinorFeatures5_HALTI4 0x00004000 #define chipMinorFeatures5_UNK15 0x00008000 -#define chipMinorFeatures5_UNK16 0x00010000 -#define chipMinorFeatures5_UNK17 0x00020000 +#define chipMinorFeatures5_ANDROID_ONLY 0x00010000 +#define chipMinorFeatures5_HAS_PRODUCTID 0x00020000 #define chipMinorFeatures5_UNK18 0x00040000 #define chipMinorFeatures5_UNK19 0x00080000 -#define chipMinorFeatures5_UNK20 0x00100000 +#define chipMinorFeatures5_PE_DITHER_FIX2 0x00100000 #define chipMinorFeatures5_UNK21 0x00200000 #define chipMinorFeatures5_UNK22 0x00400000 #define chipMinorFeatures5_UNK23 0x00800000 #define chipMinorFeatures5_UNK24 0x01000000 #define chipMinorFeatures5_UNK25 0x02000000 #define chipMinorFeatures5_UNK26 0x04000000 -#define chipMinorFeatures5_UNK27 0x08000000 -#define chipMinorFeatures5_UNK28 0x10000000 +#define chipMinorFeatures5_RS_DEPTHSTENCIL_NATIVE_SUPPORT 0x08000000 +#define chipMinorFeatures5_V2_MSAA_COMP_FIX 0x10000000 #define chipMinorFeatures5_UNK29 0x20000000 #define chipMinorFeatures5_UNK30 0x40000000 #define chipMinorFeatures5_UNK31 0x80000000 diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.c b/drivers/gpu/drm/etnaviv/etnaviv_drv.c index 5255278dde56..91e17aeee1da 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_drv.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.c @@ -495,6 +495,7 @@ static struct drm_driver etnaviv_drm_driver = { .prime_fd_to_handle = drm_gem_prime_fd_to_handle, .gem_prime_export = drm_gem_prime_export, .gem_prime_import = drm_gem_prime_import, + .gem_prime_res_obj = etnaviv_gem_prime_res_obj, .gem_prime_pin = etnaviv_gem_prime_pin, .gem_prime_unpin = etnaviv_gem_prime_unpin, .gem_prime_get_sg_table = etnaviv_gem_prime_get_sg_table, diff --git a/drivers/gpu/drm/etnaviv/etnaviv_drv.h b/drivers/gpu/drm/etnaviv/etnaviv_drv.h index e41f38667c1c..058389f93b69 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_drv.h +++ b/drivers/gpu/drm/etnaviv/etnaviv_drv.h @@ -80,6 +80,7 @@ void *etnaviv_gem_prime_vmap(struct drm_gem_object *obj); void etnaviv_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr); int etnaviv_gem_prime_mmap(struct drm_gem_object *obj, struct vm_area_struct *vma); +struct reservation_object *etnaviv_gem_prime_res_obj(struct drm_gem_object *obj); struct drm_gem_object *etnaviv_gem_prime_import_sg_table(struct drm_device *dev, struct dma_buf_attachment *attach, struct sg_table *sg); int etnaviv_gem_prime_pin(struct drm_gem_object *obj); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.c b/drivers/gpu/drm/etnaviv/etnaviv_gem.c index d6fb724fc3cc..9a3bea738330 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.c @@ -411,16 +411,20 @@ int etnaviv_gem_cpu_prep(struct drm_gem_object *obj, u32 op, struct etnaviv_gem_object *etnaviv_obj = to_etnaviv_bo(obj); struct drm_device *dev = obj->dev; bool write = !!(op & ETNA_PREP_WRITE); - unsigned long remain = - op & ETNA_PREP_NOSYNC ? 0 : etnaviv_timeout_to_jiffies(timeout); - long lret; - - lret = reservation_object_wait_timeout_rcu(etnaviv_obj->resv, - write, true, remain); - if (lret < 0) - return lret; - else if (lret == 0) - return remain == 0 ? -EBUSY : -ETIMEDOUT; + int ret; + + if (op & ETNA_PREP_NOSYNC) { + if (!reservation_object_test_signaled_rcu(etnaviv_obj->resv, + write)) + return -EBUSY; + } else { + unsigned long remain = etnaviv_timeout_to_jiffies(timeout); + + ret = reservation_object_wait_timeout_rcu(etnaviv_obj->resv, + write, true, remain); + if (ret <= 0) + return ret == 0 ? -ETIMEDOUT : ret; + } if (etnaviv_obj->flags & ETNA_BO_CACHED) { if (!etnaviv_obj->sgt) { diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem.h b/drivers/gpu/drm/etnaviv/etnaviv_gem.h index c4a091e87426..e437fba1209d 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem.h +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem.h @@ -106,9 +106,10 @@ struct etnaviv_gem_submit { struct etnaviv_gpu *gpu; struct ww_acquire_ctx ticket; struct dma_fence *fence; + u32 flags; unsigned int nr_bos; struct etnaviv_gem_submit_bo bos[0]; - u32 flags; + /* No new members here, the previous one is variable-length! */ }; int etnaviv_gem_wait_bo(struct etnaviv_gpu *gpu, struct drm_gem_object *obj, diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c index 367bf952f61a..e5da4f2300ba 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_prime.c @@ -150,3 +150,10 @@ fail: return ERR_PTR(ret); } + +struct reservation_object *etnaviv_gem_prime_res_obj(struct drm_gem_object *obj) +{ + struct etnaviv_gem_object *etnaviv_obj = to_etnaviv_bo(obj); + + return etnaviv_obj->resv; +} diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c index ee7069e93eda..5bd93169dac2 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gem_submit.c @@ -172,7 +172,7 @@ static int submit_fence_sync(const struct etnaviv_gem_submit *submit) for (i = 0; i < submit->nr_bos; i++) { struct etnaviv_gem_object *etnaviv_obj = submit->bos[i].obj; bool write = submit->bos[i].flags & ETNA_SUBMIT_BO_WRITE; - bool explicit = !(submit->flags & ETNA_SUBMIT_NO_IMPLICIT); + bool explicit = !!(submit->flags & ETNA_SUBMIT_NO_IMPLICIT); ret = etnaviv_gpu_fence_sync_obj(etnaviv_obj, context, write, explicit); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c index 9a9c40717801..ada45fdd0eae 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c @@ -412,13 +412,19 @@ static void etnaviv_gpu_load_clock(struct etnaviv_gpu *gpu, u32 clock) static void etnaviv_gpu_update_clock(struct etnaviv_gpu *gpu) { - unsigned int fscale = 1 << (6 - gpu->freq_scale); - u32 clock; - - clock = VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS | - VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(fscale); + if (gpu->identity.minor_features2 & + chipMinorFeatures2_DYNAMIC_FREQUENCY_SCALING) { + clk_set_rate(gpu->clk_core, + gpu->base_rate_core >> gpu->freq_scale); + clk_set_rate(gpu->clk_shader, + gpu->base_rate_shader >> gpu->freq_scale); + } else { + unsigned int fscale = 1 << (6 - gpu->freq_scale); + u32 clock = VIVS_HI_CLOCK_CONTROL_DISABLE_DEBUG_REGISTERS | + VIVS_HI_CLOCK_CONTROL_FSCALE_VAL(fscale); - etnaviv_gpu_load_clock(gpu, clock); + etnaviv_gpu_load_clock(gpu, clock); + } } static int etnaviv_hw_reset(struct etnaviv_gpu *gpu) @@ -523,9 +529,10 @@ static void etnaviv_gpu_enable_mlcg(struct etnaviv_gpu *gpu) pmc = gpu_read(gpu, VIVS_PM_MODULE_CONTROLS); - /* Disable PA clock gating for GC400+ except for GC420 */ + /* Disable PA clock gating for GC400+ without bugfix except for GC420 */ if (gpu->identity.model >= chipModel_GC400 && - gpu->identity.model != chipModel_GC420) + gpu->identity.model != chipModel_GC420 && + !(gpu->identity.minor_features3 & chipMinorFeatures3_BUG_FIXES12)) pmc |= VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_PA; /* @@ -541,6 +548,11 @@ static void etnaviv_gpu_enable_mlcg(struct etnaviv_gpu *gpu) if (gpu->identity.revision < 0x5422) pmc |= BIT(15); /* Unknown bit */ + /* Disable TX clock gating on affected core revisions. */ + if (etnaviv_is_model_rev(gpu, GC4000, 0x5222) || + etnaviv_is_model_rev(gpu, GC2000, 0x5108)) + pmc |= VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_TX; + pmc |= VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_RA_HZ; pmc |= VIVS_PM_MODULE_CONTROLS_DISABLE_MODULE_CLOCK_GATING_RA_EZ; @@ -1736,11 +1748,13 @@ static int etnaviv_gpu_platform_probe(struct platform_device *pdev) DBG("clk_core: %p", gpu->clk_core); if (IS_ERR(gpu->clk_core)) gpu->clk_core = NULL; + gpu->base_rate_core = clk_get_rate(gpu->clk_core); gpu->clk_shader = devm_clk_get(&pdev->dev, "shader"); DBG("clk_shader: %p", gpu->clk_shader); if (IS_ERR(gpu->clk_shader)) gpu->clk_shader = NULL; + gpu->base_rate_shader = clk_get_rate(gpu->clk_shader); /* TODO: figure out max mapped size */ dev_set_drvdata(dev, gpu); diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h index 9227a9740447..689cb8f3680c 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.h +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.h @@ -152,6 +152,8 @@ struct etnaviv_gpu { u32 hangcheck_dma_addr; struct work_struct recover_work; unsigned int freq_scale; + unsigned long base_rate_core; + unsigned long base_rate_shader; }; static inline void gpu_write(struct etnaviv_gpu *gpu, u32 reg, u32 data) diff --git a/drivers/gpu/drm/exynos/exynos_drm_fb.c b/drivers/gpu/drm/exynos/exynos_drm_fb.c index c77a5aced81a..d48fd7c918f8 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_fb.c +++ b/drivers/gpu/drm/exynos/exynos_drm_fb.c @@ -181,8 +181,8 @@ dma_addr_t exynos_drm_fb_dma_addr(struct drm_framebuffer *fb, int index) { struct exynos_drm_fb *exynos_fb = to_exynos_fb(fb); - if (index >= MAX_FB_BUFFER) - return DMA_ERROR_CODE; + if (WARN_ON_ONCE(index >= MAX_FB_BUFFER)) + return 0; return exynos_fb->dma_addr[index]; } diff --git a/drivers/gpu/drm/i915/Kconfig b/drivers/gpu/drm/i915/Kconfig index a5cd5dacf055..e9e64e8e9765 100644 --- a/drivers/gpu/drm/i915/Kconfig +++ b/drivers/gpu/drm/i915/Kconfig @@ -21,6 +21,7 @@ config DRM_I915 select ACPI_BUTTON if ACPI select SYNC_FILE select IOSF_MBI + select CRC32 help Choose this option if you have a system that has "Intel Graphics Media Accelerator" or "HD Graphics" integrated graphics, diff --git a/drivers/gpu/drm/i915/gvt/cmd_parser.c b/drivers/gpu/drm/i915/gvt/cmd_parser.c index 51241de5e7a7..713848c36349 100644 --- a/drivers/gpu/drm/i915/gvt/cmd_parser.c +++ b/drivers/gpu/drm/i915/gvt/cmd_parser.c @@ -2536,6 +2536,11 @@ static int scan_workload(struct intel_vgpu_workload *workload) gma_head == gma_tail) return 0; + if (!intel_gvt_ggtt_validate_range(s.vgpu, s.ring_start, s.ring_size)) { + ret = -EINVAL; + goto out; + } + ret = ip_gma_set(&s, gma_head); if (ret) goto out; @@ -2579,6 +2584,11 @@ static int scan_wa_ctx(struct intel_shadow_wa_ctx *wa_ctx) s.rb_va = wa_ctx->indirect_ctx.shadow_va; s.workload = workload; + if (!intel_gvt_ggtt_validate_range(s.vgpu, s.ring_start, s.ring_size)) { + ret = -EINVAL; + goto out; + } + ret = ip_gma_set(&s, gma_head); if (ret) goto out; diff --git a/drivers/gpu/drm/i915/gvt/display.c b/drivers/gpu/drm/i915/gvt/display.c index e0261fcc5b50..2deb05f618fb 100644 --- a/drivers/gpu/drm/i915/gvt/display.c +++ b/drivers/gpu/drm/i915/gvt/display.c @@ -197,6 +197,12 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | (PORT_B << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); + if (IS_BROADWELL(dev_priv)) { + vgpu_vreg(vgpu, PORT_CLK_SEL(PORT_B)) &= + ~PORT_CLK_SEL_MASK; + vgpu_vreg(vgpu, PORT_CLK_SEL(PORT_B)) |= + PORT_CLK_SEL_LCPLL_810; + } vgpu_vreg(vgpu, DDI_BUF_CTL(PORT_B)) |= DDI_BUF_CTL_ENABLE; vgpu_vreg(vgpu, DDI_BUF_CTL(PORT_B)) &= ~DDI_BUF_IS_IDLE; vgpu_vreg(vgpu, SDEISR) |= SDE_PORTB_HOTPLUG_CPT; @@ -211,6 +217,12 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | (PORT_C << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); + if (IS_BROADWELL(dev_priv)) { + vgpu_vreg(vgpu, PORT_CLK_SEL(PORT_C)) &= + ~PORT_CLK_SEL_MASK; + vgpu_vreg(vgpu, PORT_CLK_SEL(PORT_C)) |= + PORT_CLK_SEL_LCPLL_810; + } vgpu_vreg(vgpu, DDI_BUF_CTL(PORT_C)) |= DDI_BUF_CTL_ENABLE; vgpu_vreg(vgpu, DDI_BUF_CTL(PORT_C)) &= ~DDI_BUF_IS_IDLE; vgpu_vreg(vgpu, SFUSE_STRAP) |= SFUSE_STRAP_DDIC_DETECTED; @@ -225,6 +237,12 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | (PORT_D << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); + if (IS_BROADWELL(dev_priv)) { + vgpu_vreg(vgpu, PORT_CLK_SEL(PORT_D)) &= + ~PORT_CLK_SEL_MASK; + vgpu_vreg(vgpu, PORT_CLK_SEL(PORT_D)) |= + PORT_CLK_SEL_LCPLL_810; + } vgpu_vreg(vgpu, DDI_BUF_CTL(PORT_D)) |= DDI_BUF_CTL_ENABLE; vgpu_vreg(vgpu, DDI_BUF_CTL(PORT_D)) &= ~DDI_BUF_IS_IDLE; vgpu_vreg(vgpu, SFUSE_STRAP) |= SFUSE_STRAP_DDID_DETECTED; @@ -244,6 +262,10 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) vgpu_vreg(vgpu, DDI_BUF_CTL(PORT_A)) |= DDI_INIT_DISPLAY_DETECTED; } + + /* Clear host CRT status, so guest couldn't detect this host CRT. */ + if (IS_BROADWELL(dev_priv)) + vgpu_vreg(vgpu, PCH_ADPA) &= ~ADPA_CRT_HOTPLUG_MONITOR_MASK; } static void clean_virtual_dp_monitor(struct intel_vgpu *vgpu, int port_num) diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c index 66374dba3b1a..6166e34d892b 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.c +++ b/drivers/gpu/drm/i915/gvt/gtt.c @@ -2259,6 +2259,8 @@ int intel_gvt_init_gtt(struct intel_gvt *gvt) ret = setup_spt_oos(gvt); if (ret) { gvt_err("fail to initialize SPT oos\n"); + dma_unmap_page(dev, daddr, 4096, PCI_DMA_BIDIRECTIONAL); + __free_page(gvt->gtt.scratch_ggtt_page); return ret; } } diff --git a/drivers/gpu/drm/i915/gvt/handlers.c b/drivers/gpu/drm/i915/gvt/handlers.c index 1414d7e6148d..17febe830ff6 100644 --- a/drivers/gpu/drm/i915/gvt/handlers.c +++ b/drivers/gpu/drm/i915/gvt/handlers.c @@ -367,21 +367,24 @@ static int lcpll_ctl_mmio_write(struct intel_vgpu *vgpu, unsigned int offset, static int dpy_reg_mmio_read(struct intel_vgpu *vgpu, unsigned int offset, void *p_data, unsigned int bytes) { - *(u32 *)p_data = (1 << 17); - return 0; -} - -static int dpy_reg_mmio_read_2(struct intel_vgpu *vgpu, unsigned int offset, - void *p_data, unsigned int bytes) -{ - *(u32 *)p_data = 3; - return 0; -} + switch (offset) { + case 0xe651c: + case 0xe661c: + case 0xe671c: + case 0xe681c: + vgpu_vreg(vgpu, offset) = 1 << 17; + break; + case 0xe6c04: + vgpu_vreg(vgpu, offset) = 0x3; + break; + case 0xe6e1c: + vgpu_vreg(vgpu, offset) = 0x2f << 16; + break; + default: + return -EINVAL; + } -static int dpy_reg_mmio_read_3(struct intel_vgpu *vgpu, unsigned int offset, - void *p_data, unsigned int bytes) -{ - *(u32 *)p_data = (0x2f << 16); + read_vreg(vgpu, offset, p_data, bytes); return 0; } @@ -1925,7 +1928,7 @@ static int init_generic_mmio_info(struct intel_gvt *gvt) MMIO_F(_PCH_DPD_AUX_CH_CTL, 6 * 4, 0, 0, 0, D_PRE_SKL, NULL, dp_aux_ch_ctl_mmio_write); - MMIO_RO(PCH_ADPA, D_ALL, 0, ADPA_CRT_HOTPLUG_MONITOR_MASK, NULL, pch_adpa_mmio_write); + MMIO_DH(PCH_ADPA, D_PRE_SKL, NULL, pch_adpa_mmio_write); MMIO_DH(_PCH_TRANSACONF, D_ALL, NULL, transconf_mmio_write); MMIO_DH(_PCH_TRANSBCONF, D_ALL, NULL, transconf_mmio_write); @@ -2011,8 +2014,8 @@ static int init_generic_mmio_info(struct intel_gvt *gvt) MMIO_DH(0xe661c, D_ALL, dpy_reg_mmio_read, NULL); MMIO_DH(0xe671c, D_ALL, dpy_reg_mmio_read, NULL); MMIO_DH(0xe681c, D_ALL, dpy_reg_mmio_read, NULL); - MMIO_DH(0xe6c04, D_ALL, dpy_reg_mmio_read_2, NULL); - MMIO_DH(0xe6e1c, D_ALL, dpy_reg_mmio_read_3, NULL); + MMIO_DH(0xe6c04, D_ALL, dpy_reg_mmio_read, NULL); + MMIO_DH(0xe6e1c, D_ALL, dpy_reg_mmio_read, NULL); MMIO_RO(PCH_PORT_HOTPLUG, D_ALL, 0, PORTA_HOTPLUG_STATUS_MASK diff --git a/drivers/gpu/drm/i915/gvt/kvmgt.c b/drivers/gpu/drm/i915/gvt/kvmgt.c index 1ae0b4083ce1..fd0c85f9ef3c 100644 --- a/drivers/gpu/drm/i915/gvt/kvmgt.c +++ b/drivers/gpu/drm/i915/gvt/kvmgt.c @@ -232,16 +232,20 @@ static void gvt_cache_destroy(struct intel_vgpu *vgpu) struct device *dev = mdev_dev(vgpu->vdev.mdev); unsigned long gfn; - mutex_lock(&vgpu->vdev.cache_lock); - while ((node = rb_first(&vgpu->vdev.cache))) { + for (;;) { + mutex_lock(&vgpu->vdev.cache_lock); + node = rb_first(&vgpu->vdev.cache); + if (!node) { + mutex_unlock(&vgpu->vdev.cache_lock); + break; + } dma = rb_entry(node, struct gvt_dma, node); gvt_dma_unmap_iova(vgpu, dma->iova); gfn = dma->gfn; - - vfio_unpin_pages(dev, &gfn, 1); __gvt_cache_remove_entry(vgpu, dma); + mutex_unlock(&vgpu->vdev.cache_lock); + vfio_unpin_pages(dev, &gfn, 1); } - mutex_unlock(&vgpu->vdev.cache_lock); } static struct intel_vgpu_type *intel_gvt_find_vgpu_type(struct intel_gvt *gvt, diff --git a/drivers/gpu/drm/i915/gvt/scheduler.c b/drivers/gpu/drm/i915/gvt/scheduler.c index 488fdea348a9..0e2e36ad6196 100644 --- a/drivers/gpu/drm/i915/gvt/scheduler.c +++ b/drivers/gpu/drm/i915/gvt/scheduler.c @@ -174,15 +174,6 @@ static int shadow_context_status_change(struct notifier_block *nb, atomic_set(&workload->shadow_ctx_active, 1); break; case INTEL_CONTEXT_SCHEDULE_OUT: - /* If the status is -EINPROGRESS means this workload - * doesn't meet any issue during dispatching so when - * get the SCHEDULE_OUT set the status to be zero for - * good. If the status is NOT -EINPROGRESS means there - * is something wrong happened during dispatching and - * the status should not be set to zero - */ - if (workload->status == -EINPROGRESS) - workload->status = 0; atomic_set(&workload->shadow_ctx_active, 0); break; default: @@ -427,6 +418,18 @@ static void complete_current_workload(struct intel_gvt *gvt, int ring_id) wait_event(workload->shadow_ctx_status_wq, !atomic_read(&workload->shadow_ctx_active)); + /* If this request caused GPU hang, req->fence.error will + * be set to -EIO. Use -EIO to set workload status so + * that when this request caused GPU hang, didn't trigger + * context switch interrupt to guest. + */ + if (likely(workload->status == -EINPROGRESS)) { + if (workload->req->fence.error == -EIO) + workload->status = -EIO; + else + workload->status = 0; + } + i915_gem_request_put(fetch_and_zero(&workload->req)); if (!workload->status && !vgpu->resetting) { @@ -464,8 +467,6 @@ struct workload_thread_param { int ring_id; }; -static DEFINE_MUTEX(scheduler_mutex); - static int workload_thread(void *priv) { struct workload_thread_param *p = (struct workload_thread_param *)priv; @@ -497,8 +498,6 @@ static int workload_thread(void *priv) if (!workload) break; - mutex_lock(&scheduler_mutex); - gvt_dbg_sched("ring id %d next workload %p vgpu %d\n", workload->ring_id, workload, workload->vgpu->id); @@ -537,9 +536,6 @@ complete: FORCEWAKE_ALL); intel_runtime_pm_put(gvt->dev_priv); - - mutex_unlock(&scheduler_mutex); - } return 0; } @@ -620,7 +616,7 @@ err: void intel_vgpu_clean_gvt_context(struct intel_vgpu *vgpu) { - i915_gem_context_put_unlocked(vgpu->shadow_ctx); + i915_gem_context_put(vgpu->shadow_ctx); } int intel_vgpu_init_gvt_context(struct intel_vgpu *vgpu) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 4577b0af6886..2ef75c1a6119 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -292,6 +292,8 @@ static int per_file_stats(int id, void *ptr, void *data) struct file_stats *stats = data; struct i915_vma *vma; + lockdep_assert_held(&obj->base.dev->struct_mutex); + stats->count++; stats->total += obj->base.size; if (!obj->bind_count) @@ -476,6 +478,8 @@ static int i915_gem_object_info(struct seq_file *m, void *data) struct drm_i915_gem_request *request; struct task_struct *task; + mutex_lock(&dev->struct_mutex); + memset(&stats, 0, sizeof(stats)); stats.file_priv = file->driver_priv; spin_lock(&file->table_lock); @@ -487,7 +491,6 @@ static int i915_gem_object_info(struct seq_file *m, void *data) * still alive (e.g. get_pid(current) => fork() => exit()). * Therefore, we need to protect this ->comm access using RCU. */ - mutex_lock(&dev->struct_mutex); request = list_first_entry_or_null(&file_priv->mm.request_list, struct drm_i915_gem_request, client_link); @@ -497,6 +500,7 @@ static int i915_gem_object_info(struct seq_file *m, void *data) PIDTYPE_PID); print_file_stats(m, task ? task->comm : "<unknown>", stats); rcu_read_unlock(); + mutex_unlock(&dev->struct_mutex); } mutex_unlock(&dev->filelist_mutex); @@ -1155,7 +1159,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused) intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL); reqf = I915_READ(GEN6_RPNSWREQ); - if (IS_GEN9(dev_priv)) + if (INTEL_GEN(dev_priv) >= 9) reqf >>= 23; else { reqf &= ~GEN6_TURBO_DISABLE; @@ -1177,7 +1181,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused) rpdownei = I915_READ(GEN6_RP_CUR_DOWN_EI) & GEN6_CURIAVG_MASK; rpcurdown = I915_READ(GEN6_RP_CUR_DOWN) & GEN6_CURBSYTAVG_MASK; rpprevdown = I915_READ(GEN6_RP_PREV_DOWN) & GEN6_CURBSYTAVG_MASK; - if (IS_GEN9(dev_priv)) + if (INTEL_GEN(dev_priv) >= 9) cagf = (rpstat & GEN9_CAGF_MASK) >> GEN9_CAGF_SHIFT; else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) cagf = (rpstat & HSW_CAGF_MASK) >> HSW_CAGF_SHIFT; @@ -1206,7 +1210,7 @@ static int i915_frequency_info(struct seq_file *m, void *unused) dev_priv->rps.pm_intrmsk_mbz); seq_printf(m, "GT_PERF_STATUS: 0x%08x\n", gt_perf_status); seq_printf(m, "Render p-state ratio: %d\n", - (gt_perf_status & (IS_GEN9(dev_priv) ? 0x1ff00 : 0xff00)) >> 8); + (gt_perf_status & (INTEL_GEN(dev_priv) >= 9 ? 0x1ff00 : 0xff00)) >> 8); seq_printf(m, "Render p-state VID: %d\n", gt_perf_status & 0xff); seq_printf(m, "Render p-state limit: %d\n", @@ -1237,18 +1241,21 @@ static int i915_frequency_info(struct seq_file *m, void *unused) max_freq = (IS_GEN9_LP(dev_priv) ? rp_state_cap >> 0 : rp_state_cap >> 16) & 0xff; - max_freq *= (IS_GEN9_BC(dev_priv) ? GEN9_FREQ_SCALER : 1); + max_freq *= (IS_GEN9_BC(dev_priv) || + IS_CANNONLAKE(dev_priv) ? GEN9_FREQ_SCALER : 1); seq_printf(m, "Lowest (RPN) frequency: %dMHz\n", intel_gpu_freq(dev_priv, max_freq)); max_freq = (rp_state_cap & 0xff00) >> 8; - max_freq *= (IS_GEN9_BC(dev_priv) ? GEN9_FREQ_SCALER : 1); + max_freq *= (IS_GEN9_BC(dev_priv) || + IS_CANNONLAKE(dev_priv) ? GEN9_FREQ_SCALER : 1); seq_printf(m, "Nominal (RP1) frequency: %dMHz\n", intel_gpu_freq(dev_priv, max_freq)); max_freq = (IS_GEN9_LP(dev_priv) ? rp_state_cap >> 16 : rp_state_cap >> 0) & 0xff; - max_freq *= (IS_GEN9_BC(dev_priv) ? GEN9_FREQ_SCALER : 1); + max_freq *= (IS_GEN9_BC(dev_priv) || + IS_CANNONLAKE(dev_priv) ? GEN9_FREQ_SCALER : 1); seq_printf(m, "Max non-overclocked (RP0) frequency: %dMHz\n", intel_gpu_freq(dev_priv, max_freq)); seq_printf(m, "Max overclocked frequency: %dMHz\n", @@ -1403,6 +1410,23 @@ static int i915_hangcheck_info(struct seq_file *m, void *unused) return 0; } +static int i915_reset_info(struct seq_file *m, void *unused) +{ + struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct i915_gpu_error *error = &dev_priv->gpu_error; + struct intel_engine_cs *engine; + enum intel_engine_id id; + + seq_printf(m, "full gpu reset = %u\n", i915_reset_count(error)); + + for_each_engine(engine, dev_priv, id) { + seq_printf(m, "%s = %u\n", engine->name, + i915_reset_engine_count(error, engine)); + } + + return 0; +} + static int ironlake_drpc_info(struct seq_file *m) { struct drm_i915_private *dev_priv = node_to_i915(m->private); @@ -1834,7 +1858,7 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused) if (ret) goto out; - if (IS_GEN9_BC(dev_priv)) { + if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) { /* Convert GT frequency to 50 HZ units */ min_gpu_freq = dev_priv->rps.min_freq_softlimit / GEN9_FREQ_SCALER; @@ -1854,7 +1878,8 @@ static int i915_ring_freq_table(struct seq_file *m, void *unused) &ia_freq); seq_printf(m, "%d\t\t%d\t\t\t\t%d\n", intel_gpu_freq(dev_priv, (gpu_freq * - (IS_GEN9_BC(dev_priv) ? + (IS_GEN9_BC(dev_priv) || + IS_CANNONLAKE(dev_priv) ? GEN9_FREQ_SCALER : 1))), ((ia_freq >> 0) & 0xff) * 100, ((ia_freq >> 8) & 0xff) * 100); @@ -1910,7 +1935,7 @@ static int i915_gem_framebuffer_info(struct seq_file *m, void *data) return ret; #ifdef CONFIG_DRM_FBDEV_EMULATION - if (dev_priv->fbdev) { + if (dev_priv->fbdev && dev_priv->fbdev->helper.fb) { fbdev_fb = to_intel_framebuffer(dev_priv->fbdev->helper.fb); seq_printf(m, "fbcon size: %d x %d, depth %d, %d bpp, modifier 0x%llx, refcount %d, obj ", @@ -1966,7 +1991,7 @@ static int i915_context_status(struct seq_file *m, void *unused) if (ret) return ret; - list_for_each_entry(ctx, &dev_priv->context_list, link) { + list_for_each_entry(ctx, &dev_priv->contexts.list, link) { seq_printf(m, "HW context %u ", ctx->hw_id); if (ctx->pid) { struct task_struct *task; @@ -2072,7 +2097,7 @@ static int i915_dump_lrc(struct seq_file *m, void *unused) if (ret) return ret; - list_for_each_entry(ctx, &dev_priv->context_list, link) + list_for_each_entry(ctx, &dev_priv->contexts.list, link) for_each_engine(engine, dev_priv, id) i915_dump_lrc_obj(m, ctx, engine); @@ -2306,6 +2331,8 @@ static int i915_rps_boost_info(struct seq_file *m, void *data) seq_printf(m, "GPU busy? %s [%d requests]\n", yesno(dev_priv->gt.awake), dev_priv->gt.active_requests); seq_printf(m, "CPU waiting? %d\n", count_irq_waiters(dev_priv)); + seq_printf(m, "Boosts outstanding? %d\n", + atomic_read(&dev_priv->rps.num_waiters)); seq_printf(m, "Frequency requested %d\n", intel_gpu_freq(dev_priv, dev_priv->rps.cur_freq)); seq_printf(m, " min hard:%d, soft:%d; max soft:%d, hard:%d\n", @@ -2319,22 +2346,20 @@ static int i915_rps_boost_info(struct seq_file *m, void *data) intel_gpu_freq(dev_priv, dev_priv->rps.boost_freq)); mutex_lock(&dev->filelist_mutex); - spin_lock(&dev_priv->rps.client_lock); list_for_each_entry_reverse(file, &dev->filelist, lhead) { struct drm_i915_file_private *file_priv = file->driver_priv; struct task_struct *task; rcu_read_lock(); task = pid_task(file->pid, PIDTYPE_PID); - seq_printf(m, "%s [%d]: %d boosts%s\n", + seq_printf(m, "%s [%d]: %d boosts\n", task ? task->comm : "<unknown>", task ? task->pid : -1, - file_priv->rps.boosts, - list_empty(&file_priv->rps.link) ? "" : ", active"); + atomic_read(&file_priv->rps.boosts)); rcu_read_unlock(); } - seq_printf(m, "Kernel (anonymous) boosts: %d\n", dev_priv->rps.boosts); - spin_unlock(&dev_priv->rps.client_lock); + seq_printf(m, "Kernel (anonymous) boosts: %d\n", + atomic_read(&dev_priv->rps.boosts)); mutex_unlock(&dev->filelist_mutex); if (INTEL_GEN(dev_priv) >= 6 && @@ -3083,7 +3108,7 @@ static void intel_connector_info(struct seq_file *m, connector->display_info.cea_rev); } - if (!intel_encoder || intel_encoder->type == INTEL_OUTPUT_DP_MST) + if (!intel_encoder) return; switch (connector->connector_type) { @@ -3285,6 +3310,7 @@ static int i915_display_info(struct seq_file *m, void *unused) static int i915_engine_info(struct seq_file *m, void *unused) { struct drm_i915_private *dev_priv = node_to_i915(m->private); + struct i915_gpu_error *error = &dev_priv->gpu_error; struct intel_engine_cs *engine; enum intel_engine_id id; @@ -3308,6 +3334,8 @@ static int i915_engine_info(struct seq_file *m, void *unused) engine->hangcheck.seqno, jiffies_to_msecs(jiffies - engine->hangcheck.action_timestamp), engine->timeline->inflight_seqnos); + seq_printf(m, "\tReset count: %d\n", + i915_reset_engine_count(error, engine)); rcu_read_lock(); @@ -3754,13 +3782,18 @@ static ssize_t i915_displayport_test_active_write(struct file *file, drm_connector_list_iter_begin(dev, &conn_iter); drm_for_each_connector_iter(connector, &conn_iter) { + struct intel_encoder *encoder; + if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) continue; - if (connector->status == connector_status_connected && - connector->encoder != NULL) { - intel_dp = enc_to_intel_dp(connector->encoder); + encoder = to_intel_encoder(connector->encoder); + if (encoder && encoder->type == INTEL_OUTPUT_DP_MST) + continue; + + if (encoder && connector->status == connector_status_connected) { + intel_dp = enc_to_intel_dp(&encoder->base); status = kstrtoint(input_buffer, 10, &val); if (status < 0) break; @@ -3792,13 +3825,18 @@ static int i915_displayport_test_active_show(struct seq_file *m, void *data) drm_connector_list_iter_begin(dev, &conn_iter); drm_for_each_connector_iter(connector, &conn_iter) { + struct intel_encoder *encoder; + if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) continue; - if (connector->status == connector_status_connected && - connector->encoder != NULL) { - intel_dp = enc_to_intel_dp(connector->encoder); + encoder = to_intel_encoder(connector->encoder); + if (encoder && encoder->type == INTEL_OUTPUT_DP_MST) + continue; + + if (encoder && connector->status == connector_status_connected) { + intel_dp = enc_to_intel_dp(&encoder->base); if (intel_dp->compliance.test_active) seq_puts(m, "1"); else @@ -3838,13 +3876,18 @@ static int i915_displayport_test_data_show(struct seq_file *m, void *data) drm_connector_list_iter_begin(dev, &conn_iter); drm_for_each_connector_iter(connector, &conn_iter) { + struct intel_encoder *encoder; + if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) continue; - if (connector->status == connector_status_connected && - connector->encoder != NULL) { - intel_dp = enc_to_intel_dp(connector->encoder); + encoder = to_intel_encoder(connector->encoder); + if (encoder && encoder->type == INTEL_OUTPUT_DP_MST) + continue; + + if (encoder && connector->status == connector_status_connected) { + intel_dp = enc_to_intel_dp(&encoder->base); if (intel_dp->compliance.test_type == DP_TEST_LINK_EDID_READ) seq_printf(m, "%lx", @@ -3891,13 +3934,18 @@ static int i915_displayport_test_type_show(struct seq_file *m, void *data) drm_connector_list_iter_begin(dev, &conn_iter); drm_for_each_connector_iter(connector, &conn_iter) { + struct intel_encoder *encoder; + if (connector->connector_type != DRM_MODE_CONNECTOR_DisplayPort) continue; - if (connector->status == connector_status_connected && - connector->encoder != NULL) { - intel_dp = enc_to_intel_dp(connector->encoder); + encoder = to_intel_encoder(connector->encoder); + if (encoder && encoder->type == INTEL_OUTPUT_DP_MST) + continue; + + if (encoder && connector->status == connector_status_connected) { + intel_dp = enc_to_intel_dp(&encoder->base); seq_printf(m, "%02lx", intel_dp->compliance.test_type); } else seq_puts(m, "0"); @@ -4820,6 +4868,7 @@ static const struct drm_info_list i915_debugfs_list[] = { {"i915_huc_load_status", i915_huc_load_status_info, 0}, {"i915_frequency_info", i915_frequency_info, 0}, {"i915_hangcheck_info", i915_hangcheck_info, 0}, + {"i915_reset_info", i915_reset_info, 0}, {"i915_drpc_info", i915_drpc_info, 0}, {"i915_emon_status", i915_emon_status, 0}, {"i915_ring_freq_table", i915_ring_freq_table, 0}, diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c index 04d9bd84ee43..d310d8245dca 100644 --- a/drivers/gpu/drm/i915/i915_drv.c +++ b/drivers/gpu/drm/i915/i915_drv.c @@ -132,9 +132,13 @@ static enum intel_pch intel_virt_detect_pch(struct drm_i915_private *dev_priv) DRM_DEBUG_KMS("Assuming Ibex Peak PCH\n"); } else if (IS_GEN6(dev_priv) || IS_IVYBRIDGE(dev_priv)) { ret = PCH_CPT; - DRM_DEBUG_KMS("Assuming CouarPoint PCH\n"); + DRM_DEBUG_KMS("Assuming CougarPoint PCH\n"); } else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) { ret = PCH_LPT; + if (IS_HSW_ULT(dev_priv) || IS_BDW_ULT(dev_priv)) + dev_priv->pch_id = INTEL_PCH_LPT_LP_DEVICE_ID_TYPE; + else + dev_priv->pch_id = INTEL_PCH_LPT_DEVICE_ID_TYPE; DRM_DEBUG_KMS("Assuming LynxPoint PCH\n"); } else if (IS_SKYLAKE(dev_priv) || IS_KABYLAKE(dev_priv)) { ret = PCH_SPT; @@ -173,29 +177,25 @@ static void intel_detect_pch(struct drm_i915_private *dev_priv) while ((pch = pci_get_class(PCI_CLASS_BRIDGE_ISA << 8, pch))) { if (pch->vendor == PCI_VENDOR_ID_INTEL) { unsigned short id = pch->device & INTEL_PCH_DEVICE_ID_MASK; - unsigned short id_ext = pch->device & - INTEL_PCH_DEVICE_ID_MASK_EXT; + + dev_priv->pch_id = id; if (id == INTEL_PCH_IBX_DEVICE_ID_TYPE) { - dev_priv->pch_id = id; dev_priv->pch_type = PCH_IBX; DRM_DEBUG_KMS("Found Ibex Peak PCH\n"); WARN_ON(!IS_GEN5(dev_priv)); } else if (id == INTEL_PCH_CPT_DEVICE_ID_TYPE) { - dev_priv->pch_id = id; dev_priv->pch_type = PCH_CPT; DRM_DEBUG_KMS("Found CougarPoint PCH\n"); - WARN_ON(!(IS_GEN6(dev_priv) || - IS_IVYBRIDGE(dev_priv))); + WARN_ON(!IS_GEN6(dev_priv) && + !IS_IVYBRIDGE(dev_priv)); } else if (id == INTEL_PCH_PPT_DEVICE_ID_TYPE) { /* PantherPoint is CPT compatible */ - dev_priv->pch_id = id; dev_priv->pch_type = PCH_CPT; DRM_DEBUG_KMS("Found PantherPoint PCH\n"); - WARN_ON(!(IS_GEN6(dev_priv) || - IS_IVYBRIDGE(dev_priv))); + WARN_ON(!IS_GEN6(dev_priv) && + !IS_IVYBRIDGE(dev_priv)); } else if (id == INTEL_PCH_LPT_DEVICE_ID_TYPE) { - dev_priv->pch_id = id; dev_priv->pch_type = PCH_LPT; DRM_DEBUG_KMS("Found LynxPoint PCH\n"); WARN_ON(!IS_HASWELL(dev_priv) && @@ -203,51 +203,60 @@ static void intel_detect_pch(struct drm_i915_private *dev_priv) WARN_ON(IS_HSW_ULT(dev_priv) || IS_BDW_ULT(dev_priv)); } else if (id == INTEL_PCH_LPT_LP_DEVICE_ID_TYPE) { - dev_priv->pch_id = id; dev_priv->pch_type = PCH_LPT; DRM_DEBUG_KMS("Found LynxPoint LP PCH\n"); WARN_ON(!IS_HASWELL(dev_priv) && !IS_BROADWELL(dev_priv)); WARN_ON(!IS_HSW_ULT(dev_priv) && !IS_BDW_ULT(dev_priv)); + } else if (id == INTEL_PCH_WPT_DEVICE_ID_TYPE) { + /* WildcatPoint is LPT compatible */ + dev_priv->pch_type = PCH_LPT; + DRM_DEBUG_KMS("Found WildcatPoint PCH\n"); + WARN_ON(!IS_HASWELL(dev_priv) && + !IS_BROADWELL(dev_priv)); + WARN_ON(IS_HSW_ULT(dev_priv) || + IS_BDW_ULT(dev_priv)); + } else if (id == INTEL_PCH_WPT_LP_DEVICE_ID_TYPE) { + /* WildcatPoint is LPT compatible */ + dev_priv->pch_type = PCH_LPT; + DRM_DEBUG_KMS("Found WildcatPoint LP PCH\n"); + WARN_ON(!IS_HASWELL(dev_priv) && + !IS_BROADWELL(dev_priv)); + WARN_ON(!IS_HSW_ULT(dev_priv) && + !IS_BDW_ULT(dev_priv)); } else if (id == INTEL_PCH_SPT_DEVICE_ID_TYPE) { - dev_priv->pch_id = id; dev_priv->pch_type = PCH_SPT; DRM_DEBUG_KMS("Found SunrisePoint PCH\n"); WARN_ON(!IS_SKYLAKE(dev_priv) && !IS_KABYLAKE(dev_priv)); - } else if (id_ext == INTEL_PCH_SPT_LP_DEVICE_ID_TYPE) { - dev_priv->pch_id = id_ext; + } else if (id == INTEL_PCH_SPT_LP_DEVICE_ID_TYPE) { dev_priv->pch_type = PCH_SPT; DRM_DEBUG_KMS("Found SunrisePoint LP PCH\n"); WARN_ON(!IS_SKYLAKE(dev_priv) && !IS_KABYLAKE(dev_priv)); } else if (id == INTEL_PCH_KBP_DEVICE_ID_TYPE) { - dev_priv->pch_id = id; dev_priv->pch_type = PCH_KBP; DRM_DEBUG_KMS("Found KabyPoint PCH\n"); WARN_ON(!IS_SKYLAKE(dev_priv) && !IS_KABYLAKE(dev_priv)); } else if (id == INTEL_PCH_CNP_DEVICE_ID_TYPE) { - dev_priv->pch_id = id; dev_priv->pch_type = PCH_CNP; DRM_DEBUG_KMS("Found CannonPoint PCH\n"); WARN_ON(!IS_CANNONLAKE(dev_priv) && !IS_COFFEELAKE(dev_priv)); - } else if (id_ext == INTEL_PCH_CNP_LP_DEVICE_ID_TYPE) { - dev_priv->pch_id = id_ext; + } else if (id == INTEL_PCH_CNP_LP_DEVICE_ID_TYPE) { dev_priv->pch_type = PCH_CNP; DRM_DEBUG_KMS("Found CannonPoint LP PCH\n"); WARN_ON(!IS_CANNONLAKE(dev_priv) && !IS_COFFEELAKE(dev_priv)); - } else if ((id == INTEL_PCH_P2X_DEVICE_ID_TYPE) || - (id == INTEL_PCH_P3X_DEVICE_ID_TYPE) || - ((id == INTEL_PCH_QEMU_DEVICE_ID_TYPE) && + } else if (id == INTEL_PCH_P2X_DEVICE_ID_TYPE || + id == INTEL_PCH_P3X_DEVICE_ID_TYPE || + (id == INTEL_PCH_QEMU_DEVICE_ID_TYPE && pch->subsystem_vendor == PCI_SUBVENDOR_ID_REDHAT_QUMRANET && pch->subsystem_device == PCI_SUBDEVICE_ID_QEMU)) { - dev_priv->pch_id = id; dev_priv->pch_type = intel_virt_detect_pch(dev_priv); } else @@ -331,6 +340,8 @@ static int i915_getparam(struct drm_device *dev, void *data, break; case I915_PARAM_HAS_GPU_RESET: value = i915.enable_hangcheck && intel_has_gpu_reset(dev_priv); + if (value && intel_has_reset_engine(dev_priv)) + value = 2; break; case I915_PARAM_HAS_RESOURCE_STREAMER: value = HAS_RESOURCE_STREAMER(dev_priv); @@ -585,16 +596,18 @@ static const struct vga_switcheroo_client_ops i915_switcheroo_ops = { static void i915_gem_fini(struct drm_i915_private *dev_priv) { + flush_workqueue(dev_priv->wq); + mutex_lock(&dev_priv->drm.struct_mutex); intel_uc_fini_hw(dev_priv); i915_gem_cleanup_engines(dev_priv); - i915_gem_context_fini(dev_priv); + i915_gem_contexts_fini(dev_priv); i915_gem_cleanup_userptr(dev_priv); mutex_unlock(&dev_priv->drm.struct_mutex); i915_gem_drain_freed_objects(dev_priv); - WARN_ON(!list_empty(&dev_priv->context_list)); + WARN_ON(!list_empty(&dev_priv->contexts.list)); } static int i915_load_modeset_init(struct drm_device *dev) @@ -1132,10 +1145,12 @@ static int i915_driver_init_hw(struct drm_i915_private *dev_priv) * and the registers being closely associated. * * According to chipset errata, on the 965GM, MSI interrupts may - * be lost or delayed, but we use them anyways to avoid - * stuck interrupts on some machines. + * be lost or delayed, and was defeatured. MSI interrupts seem to + * get lost on g4x as well, and interrupt delivery seems to stay + * properly dead afterwards. So we'll just disable them for all + * pre-gen5 chipsets. */ - if (!IS_I945G(dev_priv) && !IS_I945GM(dev_priv)) { + if (INTEL_GEN(dev_priv) >= 5) { if (pci_enable_msi(pdev) < 0) DRM_DEBUG_DRIVER("can't enable MSI"); } @@ -1421,9 +1436,10 @@ static void i915_driver_release(struct drm_device *dev) static int i915_driver_open(struct drm_device *dev, struct drm_file *file) { + struct drm_i915_private *i915 = to_i915(dev); int ret; - ret = i915_gem_open(dev, file); + ret = i915_gem_open(i915, file); if (ret) return ret; @@ -1453,7 +1469,7 @@ static void i915_driver_postclose(struct drm_device *dev, struct drm_file *file) struct drm_i915_file_private *file_priv = file->driver_priv; mutex_lock(&dev->struct_mutex); - i915_gem_context_close(dev, file); + i915_gem_context_close(file); i915_gem_release(dev, file); mutex_unlock(&dev->struct_mutex); @@ -1905,9 +1921,72 @@ wakeup: error: i915_gem_set_wedged(dev_priv); + i915_gem_retire_requests(dev_priv); goto finish; } +/** + * i915_reset_engine - reset GPU engine to recover from a hang + * @engine: engine to reset + * + * Reset a specific GPU engine. Useful if a hang is detected. + * Returns zero on successful reset or otherwise an error code. + * + * Procedure is: + * - identifies the request that caused the hang and it is dropped + * - reset engine (which will force the engine to idle) + * - re-init/configure engine + */ +int i915_reset_engine(struct intel_engine_cs *engine) +{ + struct i915_gpu_error *error = &engine->i915->gpu_error; + struct drm_i915_gem_request *active_request; + int ret; + + GEM_BUG_ON(!test_bit(I915_RESET_ENGINE + engine->id, &error->flags)); + + DRM_DEBUG_DRIVER("resetting %s\n", engine->name); + + active_request = i915_gem_reset_prepare_engine(engine); + if (IS_ERR(active_request)) { + DRM_DEBUG_DRIVER("Previous reset failed, promote to full reset\n"); + ret = PTR_ERR(active_request); + goto out; + } + + /* + * The request that caused the hang is stuck on elsp, we know the + * active request and can drop it, adjust head to skip the offending + * request to resume executing remaining requests in the queue. + */ + i915_gem_reset_engine(engine, active_request); + + /* Finally, reset just this engine. */ + ret = intel_gpu_reset(engine->i915, intel_engine_flag(engine)); + + i915_gem_reset_finish_engine(engine); + + if (ret) { + /* If we fail here, we expect to fallback to a global reset */ + DRM_DEBUG_DRIVER("Failed to reset %s, ret=%d\n", + engine->name, ret); + goto out; + } + + /* + * The engine and its registers (and workarounds in case of render) + * have been reset to their default values. Follow the init_ring + * process to program RING_MODE, HWSP and re-enable submission. + */ + ret = engine->init_hw(engine); + if (ret) + goto out; + + error->reset_engine_count[engine->id]++; +out: + return ret; +} + static int i915_pm_suspend(struct device *kdev) { struct pci_dev *pdev = to_pci_dev(kdev); diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 2981014fcfe2..7c6fab08a2e6 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -80,8 +80,8 @@ #define DRIVER_NAME "i915" #define DRIVER_DESC "Intel Graphics" -#define DRIVER_DATE "20170619" -#define DRIVER_TIMESTAMP 1497857498 +#define DRIVER_DATE "20170717" +#define DRIVER_TIMESTAMP 1500275179 /* Use I915_STATE_WARN(x) and I915_STATE_WARN_ON() (rather than WARN() and * WARN_ON()) for hw state sanity checks to check for unexpected conditions @@ -122,7 +122,7 @@ static inline bool is_fixed16_zero(uint_fixed_16_16_t val) return false; } -static inline uint_fixed_16_16_t u32_to_fixed_16_16(uint32_t val) +static inline uint_fixed_16_16_t u32_to_fixed16(uint32_t val) { uint_fixed_16_16_t fp; @@ -132,17 +132,17 @@ static inline uint_fixed_16_16_t u32_to_fixed_16_16(uint32_t val) return fp; } -static inline uint32_t fixed_16_16_to_u32_round_up(uint_fixed_16_16_t fp) +static inline uint32_t fixed16_to_u32_round_up(uint_fixed_16_16_t fp) { return DIV_ROUND_UP(fp.val, 1 << 16); } -static inline uint32_t fixed_16_16_to_u32(uint_fixed_16_16_t fp) +static inline uint32_t fixed16_to_u32(uint_fixed_16_16_t fp) { return fp.val >> 16; } -static inline uint_fixed_16_16_t min_fixed_16_16(uint_fixed_16_16_t min1, +static inline uint_fixed_16_16_t min_fixed16(uint_fixed_16_16_t min1, uint_fixed_16_16_t min2) { uint_fixed_16_16_t min; @@ -151,7 +151,7 @@ static inline uint_fixed_16_16_t min_fixed_16_16(uint_fixed_16_16_t min1, return min; } -static inline uint_fixed_16_16_t max_fixed_16_16(uint_fixed_16_16_t max1, +static inline uint_fixed_16_16_t max_fixed16(uint_fixed_16_16_t max1, uint_fixed_16_16_t max2) { uint_fixed_16_16_t max; @@ -160,6 +160,14 @@ static inline uint_fixed_16_16_t max_fixed_16_16(uint_fixed_16_16_t max1, return max; } +static inline uint_fixed_16_16_t clamp_u64_to_fixed16(uint64_t val) +{ + uint_fixed_16_16_t fp; + WARN_ON(val >> 32); + fp.val = clamp_t(uint32_t, val, 0, ~0); + return fp; +} + static inline uint32_t div_round_up_fixed16(uint_fixed_16_16_t val, uint_fixed_16_16_t d) { @@ -170,48 +178,30 @@ static inline uint32_t mul_round_up_u32_fixed16(uint32_t val, uint_fixed_16_16_t mul) { uint64_t intermediate_val; - uint32_t result; intermediate_val = (uint64_t) val * mul.val; intermediate_val = DIV_ROUND_UP_ULL(intermediate_val, 1 << 16); WARN_ON(intermediate_val >> 32); - result = clamp_t(uint32_t, intermediate_val, 0, ~0); - return result; + return clamp_t(uint32_t, intermediate_val, 0, ~0); } static inline uint_fixed_16_16_t mul_fixed16(uint_fixed_16_16_t val, uint_fixed_16_16_t mul) { uint64_t intermediate_val; - uint_fixed_16_16_t fp; intermediate_val = (uint64_t) val.val * mul.val; intermediate_val = intermediate_val >> 16; - WARN_ON(intermediate_val >> 32); - fp.val = clamp_t(uint32_t, intermediate_val, 0, ~0); - return fp; + return clamp_u64_to_fixed16(intermediate_val); } -static inline uint_fixed_16_16_t fixed_16_16_div(uint32_t val, uint32_t d) +static inline uint_fixed_16_16_t div_fixed16(uint32_t val, uint32_t d) { - uint_fixed_16_16_t fp, res; - - fp = u32_to_fixed_16_16(val); - res.val = DIV_ROUND_UP(fp.val, d); - return res; -} - -static inline uint_fixed_16_16_t fixed_16_16_div_u64(uint32_t val, uint32_t d) -{ - uint_fixed_16_16_t res; uint64_t interm_val; interm_val = (uint64_t)val << 16; interm_val = DIV_ROUND_UP_ULL(interm_val, d); - WARN_ON(interm_val >> 32); - res.val = (uint32_t) interm_val; - - return res; + return clamp_u64_to_fixed16(interm_val); } static inline uint32_t div_round_up_u32_fixed16(uint32_t val, @@ -225,16 +215,32 @@ static inline uint32_t div_round_up_u32_fixed16(uint32_t val, return clamp_t(uint32_t, interm_val, 0, ~0); } -static inline uint_fixed_16_16_t mul_u32_fixed_16_16(uint32_t val, +static inline uint_fixed_16_16_t mul_u32_fixed16(uint32_t val, uint_fixed_16_16_t mul) { uint64_t intermediate_val; - uint_fixed_16_16_t fp; intermediate_val = (uint64_t) val * mul.val; - WARN_ON(intermediate_val >> 32); - fp.val = (uint32_t) intermediate_val; - return fp; + return clamp_u64_to_fixed16(intermediate_val); +} + +static inline uint_fixed_16_16_t add_fixed16(uint_fixed_16_16_t add1, + uint_fixed_16_16_t add2) +{ + uint64_t interm_sum; + + interm_sum = (uint64_t) add1.val + add2.val; + return clamp_u64_to_fixed16(interm_sum); +} + +static inline uint_fixed_16_16_t add_fixed16_u32(uint_fixed_16_16_t add1, + uint32_t add2) +{ + uint64_t interm_sum; + uint_fixed_16_16_t interm_add2 = u32_to_fixed16(add2); + + interm_sum = (uint64_t) add1.val + interm_add2.val; + return clamp_u64_to_fixed16(interm_sum); } static inline const char *yesno(bool v) @@ -584,8 +590,7 @@ struct drm_i915_file_private { struct idr context_idr; struct intel_rps_client { - struct list_head link; - unsigned boosts; + atomic_t boosts; } rps; unsigned int bsd_engine; @@ -753,6 +758,7 @@ struct intel_csr { func(has_csr); \ func(has_ddi); \ func(has_dp_mst); \ + func(has_reset_engine); \ func(has_fbc); \ func(has_fpga_dbg); \ func(has_full_ppgtt); \ @@ -917,6 +923,7 @@ struct i915_gpu_state { enum intel_engine_hangcheck_action hangcheck_action; struct i915_address_space *vm; int num_requests; + u32 reset_count; /* position of active request inside the ring */ u32 rq_head, rq_post, rq_tail; @@ -1149,8 +1156,8 @@ struct i915_psr { enum intel_pch { PCH_NONE = 0, /* No PCH present */ PCH_IBX, /* Ibexpeak PCH */ - PCH_CPT, /* Cougarpoint PCH */ - PCH_LPT, /* Lynxpoint PCH */ + PCH_CPT, /* Cougarpoint/Pantherpoint PCH */ + PCH_LPT, /* Lynxpoint/Wildcatpoint PCH */ PCH_SPT, /* Sunrisepoint PCH */ PCH_KBP, /* Kabypoint PCH */ PCH_CNP, /* Cannonpoint PCH */ @@ -1166,6 +1173,7 @@ enum intel_sbi_destination { #define QUIRK_INVERT_BRIGHTNESS (1<<2) #define QUIRK_BACKLIGHT_PRESENT (1<<3) #define QUIRK_PIN_SWIZZLED_PAGES (1<<5) +#define QUIRK_INCREASE_T12_DELAY (1<<6) struct intel_fbdev; struct intel_fbc_work; @@ -1301,13 +1309,10 @@ struct intel_gen6_power_mgmt { int last_adj; enum { LOW_POWER, BETWEEN, HIGH_POWER } power; - spinlock_t client_lock; - struct list_head clients; - bool client_boost; - bool enabled; struct delayed_work autoenable_work; - unsigned boosts; + atomic_t num_waiters; + atomic_t boosts; /* manual wa residency calculations */ struct intel_rps_ei ei; @@ -1550,6 +1555,12 @@ struct i915_gpu_error { * inspect the bit and do the reset directly, otherwise the worker * waits for the struct_mutex. * + * #I915_RESET_ENGINE[num_engines] - Since the driver doesn't need to + * acquire the struct_mutex to reset an engine, we need an explicit + * flag to prevent two concurrent reset attempts in the same engine. + * As the number of engines continues to grow, allocate the flags from + * the most significant bits. + * * #I915_WEDGED - If reset fails and we can no longer use the GPU, * we set the #I915_WEDGED bit. Prior to command submission, e.g. * i915_gem_request_alloc(), this bit is checked and the sequence @@ -1559,6 +1570,10 @@ struct i915_gpu_error { #define I915_RESET_BACKOFF 0 #define I915_RESET_HANDOFF 1 #define I915_WEDGED (BITS_PER_LONG - 1) +#define I915_RESET_ENGINE (I915_WEDGED - I915_NUM_ENGINES) + + /** Number of times an engine has been reset */ + u32 reset_engine_count[I915_NUM_ENGINES]; /** * Waitqueue to signal when a hang is detected. Used to for waiters @@ -2236,13 +2251,6 @@ struct drm_i915_private { DECLARE_HASHTABLE(mm_structs, 7); struct mutex mm_lock; - /* The hw wants to have a stable context identifier for the lifetime - * of the context (for OA, PASID, faults, etc). This is limited - * in execlists to 21 bits. - */ - struct ida context_hw_ida; -#define MAX_CONTEXT_HW_ID (1<<21) /* exclusive */ - /* Kernel Modesetting */ struct intel_crtc *plane_to_crtc_mapping[I915_MAX_PIPES]; @@ -2319,7 +2327,18 @@ struct drm_i915_private { */ struct mutex av_mutex; - struct list_head context_list; + struct { + struct list_head list; + struct llist_head free_list; + struct work_struct free_work; + + /* The hw wants to have a stable context identifier for the + * lifetime of the context (for OA, PASID, faults, etc). + * This is limited in execlists to 21 bits. + */ + struct ida hw_ida; +#define MAX_CONTEXT_HW_ID (1<<21) /* exclusive */ + } contexts; u32 fdi_rx_config; @@ -2994,16 +3013,17 @@ intel_info(const struct drm_i915_private *dev_priv) #define HAS_POOLED_EU(dev_priv) ((dev_priv)->info.has_pooled_eu) -#define INTEL_PCH_DEVICE_ID_MASK 0xff00 -#define INTEL_PCH_DEVICE_ID_MASK_EXT 0xff80 +#define INTEL_PCH_DEVICE_ID_MASK 0xff80 #define INTEL_PCH_IBX_DEVICE_ID_TYPE 0x3b00 #define INTEL_PCH_CPT_DEVICE_ID_TYPE 0x1c00 #define INTEL_PCH_PPT_DEVICE_ID_TYPE 0x1e00 #define INTEL_PCH_LPT_DEVICE_ID_TYPE 0x8c00 #define INTEL_PCH_LPT_LP_DEVICE_ID_TYPE 0x9c00 +#define INTEL_PCH_WPT_DEVICE_ID_TYPE 0x8c80 +#define INTEL_PCH_WPT_LP_DEVICE_ID_TYPE 0x9c80 #define INTEL_PCH_SPT_DEVICE_ID_TYPE 0xA100 #define INTEL_PCH_SPT_LP_DEVICE_ID_TYPE 0x9D00 -#define INTEL_PCH_KBP_DEVICE_ID_TYPE 0xA200 +#define INTEL_PCH_KBP_DEVICE_ID_TYPE 0xA280 #define INTEL_PCH_CNP_DEVICE_ID_TYPE 0xA300 #define INTEL_PCH_CNP_LP_DEVICE_ID_TYPE 0x9D80 #define INTEL_PCH_P2X_DEVICE_ID_TYPE 0x7100 @@ -3018,9 +3038,11 @@ intel_info(const struct drm_i915_private *dev_priv) #define HAS_PCH_SPT(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_SPT) #define HAS_PCH_LPT(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_LPT) #define HAS_PCH_LPT_LP(dev_priv) \ - ((dev_priv)->pch_id == INTEL_PCH_LPT_LP_DEVICE_ID_TYPE) + ((dev_priv)->pch_id == INTEL_PCH_LPT_LP_DEVICE_ID_TYPE || \ + (dev_priv)->pch_id == INTEL_PCH_WPT_LP_DEVICE_ID_TYPE) #define HAS_PCH_LPT_H(dev_priv) \ - ((dev_priv)->pch_id == INTEL_PCH_LPT_DEVICE_ID_TYPE) + ((dev_priv)->pch_id == INTEL_PCH_LPT_DEVICE_ID_TYPE || \ + (dev_priv)->pch_id == INTEL_PCH_WPT_DEVICE_ID_TYPE) #define HAS_PCH_CPT(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_CPT) #define HAS_PCH_IBX(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_IBX) #define HAS_PCH_NOP(dev_priv) (INTEL_PCH_TYPE(dev_priv) == PCH_NOP) @@ -3087,6 +3109,8 @@ extern void i915_driver_unload(struct drm_device *dev); extern int intel_gpu_reset(struct drm_i915_private *dev_priv, u32 engine_mask); extern bool intel_has_gpu_reset(struct drm_i915_private *dev_priv); extern void i915_reset(struct drm_i915_private *dev_priv); +extern int i915_reset_engine(struct intel_engine_cs *engine); +extern bool intel_has_reset_engine(struct drm_i915_private *dev_priv); extern int intel_guc_reset(struct drm_i915_private *dev_priv); extern void intel_engine_init_hangcheck(struct intel_engine_cs *engine); extern void intel_hangcheck_init(struct drm_i915_private *dev_priv); @@ -3459,11 +3483,22 @@ static inline u32 i915_reset_count(struct i915_gpu_error *error) return READ_ONCE(error->reset_count); } +static inline u32 i915_reset_engine_count(struct i915_gpu_error *error, + struct intel_engine_cs *engine) +{ + return READ_ONCE(error->reset_engine_count[engine->id]); +} + +struct drm_i915_gem_request * +i915_gem_reset_prepare_engine(struct intel_engine_cs *engine); int i915_gem_reset_prepare(struct drm_i915_private *dev_priv); void i915_gem_reset(struct drm_i915_private *dev_priv); +void i915_gem_reset_finish_engine(struct intel_engine_cs *engine); void i915_gem_reset_finish(struct drm_i915_private *dev_priv); void i915_gem_set_wedged(struct drm_i915_private *dev_priv); bool i915_gem_unset_wedged(struct drm_i915_private *dev_priv); +void i915_gem_reset_engine(struct intel_engine_cs *engine, + struct drm_i915_gem_request *request); void i915_gem_init_mmio(struct drm_i915_private *i915); int __must_check i915_gem_init(struct drm_i915_private *dev_priv); @@ -3497,7 +3532,7 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj, void i915_gem_object_unpin_from_display_plane(struct i915_vma *vma); int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align); -int i915_gem_open(struct drm_device *dev, struct drm_file *file); +int i915_gem_open(struct drm_i915_private *i915, struct drm_file *file); void i915_gem_release(struct drm_device *dev, struct drm_file *file); int i915_gem_object_set_cache_level(struct drm_i915_gem_object *obj, @@ -3529,38 +3564,23 @@ void i915_gem_object_save_bit_17_swizzle(struct drm_i915_gem_object *obj, struct sg_table *pages); static inline struct i915_gem_context * -i915_gem_context_lookup(struct drm_i915_file_private *file_priv, u32 id) +__i915_gem_context_lookup_rcu(struct drm_i915_file_private *file_priv, u32 id) { - struct i915_gem_context *ctx; - - lockdep_assert_held(&file_priv->dev_priv->drm.struct_mutex); - - ctx = idr_find(&file_priv->context_idr, id); - if (!ctx) - return ERR_PTR(-ENOENT); - - return ctx; + return idr_find(&file_priv->context_idr, id); } static inline struct i915_gem_context * -i915_gem_context_get(struct i915_gem_context *ctx) -{ - kref_get(&ctx->ref); - return ctx; -} - -static inline void i915_gem_context_put(struct i915_gem_context *ctx) +i915_gem_context_lookup(struct drm_i915_file_private *file_priv, u32 id) { - lockdep_assert_held(&ctx->i915->drm.struct_mutex); - kref_put(&ctx->ref, i915_gem_context_free); -} + struct i915_gem_context *ctx; -static inline void i915_gem_context_put_unlocked(struct i915_gem_context *ctx) -{ - struct mutex *lock = &ctx->i915->drm.struct_mutex; + rcu_read_lock(); + ctx = __i915_gem_context_lookup_rcu(file_priv, id); + if (ctx && !kref_get_unless_zero(&ctx->ref)) + ctx = NULL; + rcu_read_unlock(); - if (kref_put_mutex(&ctx->ref, i915_gem_context_free, lock)) - mutex_unlock(lock); + return ctx; } static inline struct intel_timeline * diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 7dcac3bfb771..d6f9b4cb6e9b 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -388,7 +388,7 @@ i915_gem_object_wait_fence(struct dma_fence *fence, */ if (rps) { if (INTEL_GEN(rq->i915) >= 6) - gen6_rps_boost(rq->i915, rps, rq->emitted_jiffies); + gen6_rps_boost(rq, rps); else rps = NULL; } @@ -399,22 +399,6 @@ out: if (flags & I915_WAIT_LOCKED && i915_gem_request_completed(rq)) i915_gem_request_retire_upto(rq); - if (rps && i915_gem_request_global_seqno(rq) == intel_engine_last_submit(rq->engine)) { - /* The GPU is now idle and this client has stalled. - * Since no other client has submitted a request in the - * meantime, assume that this client is the only one - * supplying work to the GPU but is unable to keep that - * work supplied because it is waiting. Since the GPU is - * then never kept fully busy, RPS autoclocking will - * keep the clocks relatively low, causing further delays. - * Compensate by giving the synchronous client credit for - * a waitboost next time. - */ - spin_lock(&rq->i915->rps.client_lock); - list_del_init(&rps->link); - spin_unlock(&rq->i915->rps.client_lock); - } - return timeout; } @@ -2434,8 +2418,9 @@ rebuild_st: * again with !__GFP_NORETRY. However, we still * want to fail this allocation rather than * trigger the out-of-memory killer and for - * this we want the future __GFP_MAYFAIL. + * this we want __GFP_RETRY_MAYFAIL. */ + gfp |= __GFP_RETRY_MAYFAIL; } } while (1); @@ -2831,46 +2816,64 @@ static bool engine_stalled(struct intel_engine_cs *engine) return true; } +/* + * Ensure irq handler finishes, and not run again. + * Also return the active request so that we only search for it once. + */ +struct drm_i915_gem_request * +i915_gem_reset_prepare_engine(struct intel_engine_cs *engine) +{ + struct drm_i915_gem_request *request = NULL; + + /* Prevent the signaler thread from updating the request + * state (by calling dma_fence_signal) as we are processing + * the reset. The write from the GPU of the seqno is + * asynchronous and the signaler thread may see a different + * value to us and declare the request complete, even though + * the reset routine have picked that request as the active + * (incomplete) request. This conflict is not handled + * gracefully! + */ + kthread_park(engine->breadcrumbs.signaler); + + /* Prevent request submission to the hardware until we have + * completed the reset in i915_gem_reset_finish(). If a request + * is completed by one engine, it may then queue a request + * to a second via its engine->irq_tasklet *just* as we are + * calling engine->init_hw() and also writing the ELSP. + * Turning off the engine->irq_tasklet until the reset is over + * prevents the race. + */ + tasklet_kill(&engine->irq_tasklet); + tasklet_disable(&engine->irq_tasklet); + + if (engine->irq_seqno_barrier) + engine->irq_seqno_barrier(engine); + + if (engine_stalled(engine)) { + request = i915_gem_find_active_request(engine); + if (request && request->fence.error == -EIO) + request = ERR_PTR(-EIO); /* Previous reset failed! */ + } + + return request; +} + int i915_gem_reset_prepare(struct drm_i915_private *dev_priv) { struct intel_engine_cs *engine; + struct drm_i915_gem_request *request; enum intel_engine_id id; int err = 0; - /* Ensure irq handler finishes, and not run again. */ for_each_engine(engine, dev_priv, id) { - struct drm_i915_gem_request *request; - - /* Prevent the signaler thread from updating the request - * state (by calling dma_fence_signal) as we are processing - * the reset. The write from the GPU of the seqno is - * asynchronous and the signaler thread may see a different - * value to us and declare the request complete, even though - * the reset routine have picked that request as the active - * (incomplete) request. This conflict is not handled - * gracefully! - */ - kthread_park(engine->breadcrumbs.signaler); - - /* Prevent request submission to the hardware until we have - * completed the reset in i915_gem_reset_finish(). If a request - * is completed by one engine, it may then queue a request - * to a second via its engine->irq_tasklet *just* as we are - * calling engine->init_hw() and also writing the ELSP. - * Turning off the engine->irq_tasklet until the reset is over - * prevents the race. - */ - tasklet_kill(&engine->irq_tasklet); - tasklet_disable(&engine->irq_tasklet); - - if (engine->irq_seqno_barrier) - engine->irq_seqno_barrier(engine); - - if (engine_stalled(engine)) { - request = i915_gem_find_active_request(engine); - if (request && request->fence.error == -EIO) - err = -EIO; /* Previous reset failed! */ + request = i915_gem_reset_prepare_engine(engine); + if (IS_ERR(request)) { + err = PTR_ERR(request); + continue; } + + engine->hangcheck.active_request = request; } i915_gem_revoke_fences(dev_priv); @@ -2924,7 +2927,7 @@ static void engine_skip_context(struct drm_i915_gem_request *request) static bool i915_gem_reset_request(struct drm_i915_gem_request *request) { /* Read once and return the resolution */ - const bool guilty = engine_stalled(request->engine); + const bool guilty = !i915_gem_request_completed(request); /* The guilty request will get skipped on a hung engine. * @@ -2958,11 +2961,9 @@ static bool i915_gem_reset_request(struct drm_i915_gem_request *request) return guilty; } -static void i915_gem_reset_engine(struct intel_engine_cs *engine) +void i915_gem_reset_engine(struct intel_engine_cs *engine, + struct drm_i915_gem_request *request) { - struct drm_i915_gem_request *request; - - request = i915_gem_find_active_request(engine); if (request && i915_gem_reset_request(request)) { DRM_DEBUG_DRIVER("resetting %s to restart from tail of request 0x%x\n", engine->name, request->global_seqno); @@ -2988,7 +2989,7 @@ void i915_gem_reset(struct drm_i915_private *dev_priv) for_each_engine(engine, dev_priv, id) { struct i915_gem_context *ctx; - i915_gem_reset_engine(engine); + i915_gem_reset_engine(engine, engine->hangcheck.active_request); ctx = fetch_and_zero(&engine->last_retired_context); if (ctx) engine->context_unpin(engine, ctx); @@ -3004,6 +3005,12 @@ void i915_gem_reset(struct drm_i915_private *dev_priv) } } +void i915_gem_reset_finish_engine(struct intel_engine_cs *engine) +{ + tasklet_enable(&engine->irq_tasklet); + kthread_unpark(engine->breadcrumbs.signaler); +} + void i915_gem_reset_finish(struct drm_i915_private *dev_priv) { struct intel_engine_cs *engine; @@ -3012,8 +3019,8 @@ void i915_gem_reset_finish(struct drm_i915_private *dev_priv) lockdep_assert_held(&dev_priv->drm.struct_mutex); for_each_engine(engine, dev_priv, id) { - tasklet_enable(&engine->irq_tasklet); - kthread_unpark(engine->breadcrumbs.signaler); + engine->hangcheck.active_request = NULL; + i915_gem_reset_finish_engine(engine); } } @@ -3040,7 +3047,8 @@ static void engine_set_wedged(struct intel_engine_cs *engine) /* Mark all executing requests as skipped */ spin_lock_irqsave(&engine->timeline->lock, flags); list_for_each_entry(request, &engine->timeline->requests, link) - dma_fence_set_error(&request->fence, -EIO); + if (!i915_gem_request_completed(request)) + dma_fence_set_error(&request->fence, -EIO); spin_unlock_irqrestore(&engine->timeline->lock, flags); /* Mark all pending requests as complete so that any concurrent @@ -3070,6 +3078,13 @@ static void engine_set_wedged(struct intel_engine_cs *engine) engine->execlist_first = NULL; spin_unlock_irqrestore(&engine->timeline->lock, flags); + + /* The port is checked prior to scheduling a tasklet, but + * just in case we have suspended the tasklet to do the + * wedging make sure that when it wakes, it decides there + * is no work to do by clearing the irq_posted bit. + */ + clear_bit(ENGINE_IRQ_EXECLIST, &engine->irq_posted); } } @@ -3079,6 +3094,7 @@ static int __i915_gem_set_wedged_BKL(void *data) struct intel_engine_cs *engine; enum intel_engine_id id; + set_bit(I915_WEDGED, &i915->gpu_error.flags); for_each_engine(engine, i915, id) engine_set_wedged(engine); @@ -3087,20 +3103,7 @@ static int __i915_gem_set_wedged_BKL(void *data) void i915_gem_set_wedged(struct drm_i915_private *dev_priv) { - lockdep_assert_held(&dev_priv->drm.struct_mutex); - set_bit(I915_WEDGED, &dev_priv->gpu_error.flags); - - /* Retire completed requests first so the list of inflight/incomplete - * requests is accurate and we don't try and mark successful requests - * as in error during __i915_gem_set_wedged_BKL(). - */ - i915_gem_retire_requests(dev_priv); - stop_machine(__i915_gem_set_wedged_BKL, dev_priv, NULL); - - i915_gem_context_lost(dev_priv); - - mod_delayed_work(dev_priv->wq, &dev_priv->gt.idle_work, 0); } bool i915_gem_unset_wedged(struct drm_i915_private *i915) @@ -3155,6 +3158,7 @@ bool i915_gem_unset_wedged(struct drm_i915_private *i915) * context and do not require stop_machine(). */ intel_engines_reset_default_submission(i915); + i915_gem_contexts_lost(i915); smp_mb__before_atomic(); /* complete takeover before enabling execbuf */ clear_bit(I915_WEDGED, &i915->gpu_error.flags); @@ -4564,7 +4568,7 @@ int i915_gem_suspend(struct drm_i915_private *dev_priv) goto err_unlock; assert_kernel_context_is_current(dev_priv); - i915_gem_context_lost(dev_priv); + i915_gem_contexts_lost(dev_priv); mutex_unlock(&dev->struct_mutex); intel_guc_suspend(dev_priv); @@ -4578,8 +4582,6 @@ int i915_gem_suspend(struct drm_i915_private *dev_priv) while (flush_delayed_work(&dev_priv->gt.idle_work)) ; - i915_gem_drain_freed_objects(dev_priv); - /* Assert that we sucessfully flushed all the work and * reset the GPU back to its idle, low power state. */ @@ -4811,7 +4813,7 @@ int i915_gem_init(struct drm_i915_private *dev_priv) if (ret) goto out_unlock; - ret = i915_gem_context_init(dev_priv); + ret = i915_gem_contexts_init(dev_priv); if (ret) goto out_unlock; @@ -4921,7 +4923,6 @@ i915_gem_load_init(struct drm_i915_private *dev_priv) if (err) goto err_priorities; - INIT_LIST_HEAD(&dev_priv->context_list); INIT_WORK(&dev_priv->mm.free_work, __i915_gem_free_work); init_llist_head(&dev_priv->mm.free_list); INIT_LIST_HEAD(&dev_priv->mm.unbound_list); @@ -5037,15 +5038,9 @@ void i915_gem_release(struct drm_device *dev, struct drm_file *file) list_for_each_entry(request, &file_priv->mm.request_list, client_link) request->file_priv = NULL; spin_unlock(&file_priv->mm.lock); - - if (!list_empty(&file_priv->rps.link)) { - spin_lock(&to_i915(dev)->rps.client_lock); - list_del(&file_priv->rps.link); - spin_unlock(&to_i915(dev)->rps.client_lock); - } } -int i915_gem_open(struct drm_device *dev, struct drm_file *file) +int i915_gem_open(struct drm_i915_private *i915, struct drm_file *file) { struct drm_i915_file_private *file_priv; int ret; @@ -5057,16 +5052,15 @@ int i915_gem_open(struct drm_device *dev, struct drm_file *file) return -ENOMEM; file->driver_priv = file_priv; - file_priv->dev_priv = to_i915(dev); + file_priv->dev_priv = i915; file_priv->file = file; - INIT_LIST_HEAD(&file_priv->rps.link); spin_lock_init(&file_priv->mm.lock); INIT_LIST_HEAD(&file_priv->mm.request_list); file_priv->bsd_engine = -1; - ret = i915_gem_context_open(dev, file); + ret = i915_gem_context_open(i915, file); if (ret) kfree(file_priv); diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c index 39ed58a21fc1..1a87d04e7937 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.c +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -158,13 +158,11 @@ static void vma_lut_free(struct i915_gem_context *ctx) kvfree(lut->ht); } -void i915_gem_context_free(struct kref *ctx_ref) +static void i915_gem_context_free(struct i915_gem_context *ctx) { - struct i915_gem_context *ctx = container_of(ctx_ref, typeof(*ctx), ref); int i; lockdep_assert_held(&ctx->i915->drm.struct_mutex); - trace_i915_context_free(ctx); GEM_BUG_ON(!i915_gem_context_is_closed(ctx)); vma_lut_free(ctx); @@ -188,8 +186,54 @@ void i915_gem_context_free(struct kref *ctx_ref) list_del(&ctx->link); - ida_simple_remove(&ctx->i915->context_hw_ida, ctx->hw_id); - kfree(ctx); + ida_simple_remove(&ctx->i915->contexts.hw_ida, ctx->hw_id); + kfree_rcu(ctx, rcu); +} + +static void contexts_free(struct drm_i915_private *i915) +{ + struct llist_node *freed = llist_del_all(&i915->contexts.free_list); + struct i915_gem_context *ctx, *cn; + + lockdep_assert_held(&i915->drm.struct_mutex); + + llist_for_each_entry_safe(ctx, cn, freed, free_link) + i915_gem_context_free(ctx); +} + +static void contexts_free_first(struct drm_i915_private *i915) +{ + struct i915_gem_context *ctx; + struct llist_node *freed; + + lockdep_assert_held(&i915->drm.struct_mutex); + + freed = llist_del_first(&i915->contexts.free_list); + if (!freed) + return; + + ctx = container_of(freed, typeof(*ctx), free_link); + i915_gem_context_free(ctx); +} + +static void contexts_free_worker(struct work_struct *work) +{ + struct drm_i915_private *i915 = + container_of(work, typeof(*i915), contexts.free_work); + + mutex_lock(&i915->drm.struct_mutex); + contexts_free(i915); + mutex_unlock(&i915->drm.struct_mutex); +} + +void i915_gem_context_release(struct kref *ref) +{ + struct i915_gem_context *ctx = container_of(ref, typeof(*ctx), ref); + struct drm_i915_private *i915 = ctx->i915; + + trace_i915_context_free(ctx); + if (llist_add(&ctx->free_link, &i915->contexts.free_list)) + queue_work(i915->wq, &i915->contexts.free_work); } static void context_close(struct i915_gem_context *ctx) @@ -205,7 +249,7 @@ static int assign_hw_id(struct drm_i915_private *dev_priv, unsigned *out) { int ret; - ret = ida_simple_get(&dev_priv->context_hw_ida, + ret = ida_simple_get(&dev_priv->contexts.hw_ida, 0, MAX_CONTEXT_HW_ID, GFP_KERNEL); if (ret < 0) { /* Contexts are only released when no longer active. @@ -213,7 +257,7 @@ static int assign_hw_id(struct drm_i915_private *dev_priv, unsigned *out) * stale contexts and try again. */ i915_gem_retire_requests(dev_priv); - ret = ida_simple_get(&dev_priv->context_hw_ida, + ret = ida_simple_get(&dev_priv->contexts.hw_ida, 0, MAX_CONTEXT_HW_ID, GFP_KERNEL); if (ret < 0) return ret; @@ -265,7 +309,7 @@ __create_hw_context(struct drm_i915_private *dev_priv, } kref_init(&ctx->ref); - list_add_tail(&ctx->link, &dev_priv->context_list); + list_add_tail(&ctx->link, &dev_priv->contexts.list); ctx->i915 = dev_priv; ctx->priority = I915_PRIORITY_NORMAL; @@ -354,6 +398,9 @@ i915_gem_create_context(struct drm_i915_private *dev_priv, lockdep_assert_held(&dev_priv->drm.struct_mutex); + /* Reap the most stale context */ + contexts_free_first(dev_priv); + ctx = __create_hw_context(dev_priv, file_priv); if (IS_ERR(ctx)) return ctx; @@ -418,7 +465,7 @@ out: return ctx; } -int i915_gem_context_init(struct drm_i915_private *dev_priv) +int i915_gem_contexts_init(struct drm_i915_private *dev_priv) { struct i915_gem_context *ctx; @@ -427,6 +474,10 @@ int i915_gem_context_init(struct drm_i915_private *dev_priv) if (WARN_ON(dev_priv->kernel_context)) return 0; + INIT_LIST_HEAD(&dev_priv->contexts.list); + INIT_WORK(&dev_priv->contexts.free_work, contexts_free_worker); + init_llist_head(&dev_priv->contexts.free_list); + if (intel_vgpu_active(dev_priv) && HAS_LOGICAL_RING_CONTEXTS(dev_priv)) { if (!i915.enable_execlists) { @@ -437,7 +488,7 @@ int i915_gem_context_init(struct drm_i915_private *dev_priv) /* Using the simple ida interface, the max is limited by sizeof(int) */ BUILD_BUG_ON(MAX_CONTEXT_HW_ID > INT_MAX); - ida_init(&dev_priv->context_hw_ida); + ida_init(&dev_priv->contexts.hw_ida); ctx = i915_gem_create_context(dev_priv, NULL); if (IS_ERR(ctx)) { @@ -463,7 +514,7 @@ int i915_gem_context_init(struct drm_i915_private *dev_priv) return 0; } -void i915_gem_context_lost(struct drm_i915_private *dev_priv) +void i915_gem_contexts_lost(struct drm_i915_private *dev_priv) { struct intel_engine_cs *engine; enum intel_engine_id id; @@ -484,7 +535,7 @@ void i915_gem_context_lost(struct drm_i915_private *dev_priv) if (!i915.enable_execlists) { struct i915_gem_context *ctx; - list_for_each_entry(ctx, &dev_priv->context_list, link) { + list_for_each_entry(ctx, &dev_priv->contexts.list, link) { if (!i915_gem_context_is_default(ctx)) continue; @@ -503,18 +554,20 @@ void i915_gem_context_lost(struct drm_i915_private *dev_priv) } } -void i915_gem_context_fini(struct drm_i915_private *dev_priv) +void i915_gem_contexts_fini(struct drm_i915_private *i915) { - struct i915_gem_context *dctx = dev_priv->kernel_context; - - lockdep_assert_held(&dev_priv->drm.struct_mutex); + struct i915_gem_context *ctx; - GEM_BUG_ON(!i915_gem_context_is_kernel(dctx)); + lockdep_assert_held(&i915->drm.struct_mutex); - context_close(dctx); - dev_priv->kernel_context = NULL; + /* Keep the context so that we can free it immediately ourselves */ + ctx = i915_gem_context_get(fetch_and_zero(&i915->kernel_context)); + GEM_BUG_ON(!i915_gem_context_is_kernel(ctx)); + context_close(ctx); + i915_gem_context_free(ctx); - ida_destroy(&dev_priv->context_hw_ida); + /* Must free all deferred contexts (via flush_workqueue) first */ + ida_destroy(&i915->contexts.hw_ida); } static int context_idr_cleanup(int id, void *p, void *data) @@ -525,32 +578,32 @@ static int context_idr_cleanup(int id, void *p, void *data) return 0; } -int i915_gem_context_open(struct drm_device *dev, struct drm_file *file) +int i915_gem_context_open(struct drm_i915_private *i915, + struct drm_file *file) { struct drm_i915_file_private *file_priv = file->driver_priv; struct i915_gem_context *ctx; idr_init(&file_priv->context_idr); - mutex_lock(&dev->struct_mutex); - ctx = i915_gem_create_context(to_i915(dev), file_priv); - mutex_unlock(&dev->struct_mutex); - - GEM_BUG_ON(i915_gem_context_is_kernel(ctx)); - + mutex_lock(&i915->drm.struct_mutex); + ctx = i915_gem_create_context(i915, file_priv); + mutex_unlock(&i915->drm.struct_mutex); if (IS_ERR(ctx)) { idr_destroy(&file_priv->context_idr); return PTR_ERR(ctx); } + GEM_BUG_ON(i915_gem_context_is_kernel(ctx)); + return 0; } -void i915_gem_context_close(struct drm_device *dev, struct drm_file *file) +void i915_gem_context_close(struct drm_file *file) { struct drm_i915_file_private *file_priv = file->driver_priv; - lockdep_assert_held(&dev->struct_mutex); + lockdep_assert_held(&file_priv->dev_priv->drm.struct_mutex); idr_for_each(&file_priv->context_idr, context_idr_cleanup, NULL); idr_destroy(&file_priv->context_idr); @@ -981,20 +1034,19 @@ int i915_gem_context_destroy_ioctl(struct drm_device *dev, void *data, if (args->ctx_id == DEFAULT_CONTEXT_HANDLE) return -ENOENT; - ret = i915_mutex_lock_interruptible(dev); - if (ret) - return ret; - ctx = i915_gem_context_lookup(file_priv, args->ctx_id); - if (IS_ERR(ctx)) { - mutex_unlock(&dev->struct_mutex); - return PTR_ERR(ctx); - } + if (!ctx) + return -ENOENT; + + ret = mutex_lock_interruptible(&dev->struct_mutex); + if (ret) + goto out; __destroy_hw_context(ctx, file_priv); mutex_unlock(&dev->struct_mutex); - DRM_DEBUG("HW context %d destroyed\n", args->ctx_id); +out: + i915_gem_context_put(ctx); return 0; } @@ -1004,17 +1056,11 @@ int i915_gem_context_getparam_ioctl(struct drm_device *dev, void *data, struct drm_i915_file_private *file_priv = file->driver_priv; struct drm_i915_gem_context_param *args = data; struct i915_gem_context *ctx; - int ret; - - ret = i915_mutex_lock_interruptible(dev); - if (ret) - return ret; + int ret = 0; ctx = i915_gem_context_lookup(file_priv, args->ctx_id); - if (IS_ERR(ctx)) { - mutex_unlock(&dev->struct_mutex); - return PTR_ERR(ctx); - } + if (!ctx) + return -ENOENT; args->size = 0; switch (args->param) { @@ -1042,8 +1088,8 @@ int i915_gem_context_getparam_ioctl(struct drm_device *dev, void *data, ret = -EINVAL; break; } - mutex_unlock(&dev->struct_mutex); + i915_gem_context_put(ctx); return ret; } @@ -1055,15 +1101,13 @@ int i915_gem_context_setparam_ioctl(struct drm_device *dev, void *data, struct i915_gem_context *ctx; int ret; + ctx = i915_gem_context_lookup(file_priv, args->ctx_id); + if (!ctx) + return -ENOENT; + ret = i915_mutex_lock_interruptible(dev); if (ret) - return ret; - - ctx = i915_gem_context_lookup(file_priv, args->ctx_id); - if (IS_ERR(ctx)) { - mutex_unlock(&dev->struct_mutex); - return PTR_ERR(ctx); - } + goto out; switch (args->param) { case I915_CONTEXT_PARAM_BAN_PERIOD: @@ -1101,6 +1145,8 @@ int i915_gem_context_setparam_ioctl(struct drm_device *dev, void *data, } mutex_unlock(&dev->struct_mutex); +out: + i915_gem_context_put(ctx); return ret; } @@ -1115,27 +1161,31 @@ int i915_gem_context_reset_stats_ioctl(struct drm_device *dev, if (args->flags || args->pad) return -EINVAL; - ret = i915_mutex_lock_interruptible(dev); - if (ret) - return ret; + ret = -ENOENT; + rcu_read_lock(); + ctx = __i915_gem_context_lookup_rcu(file->driver_priv, args->ctx_id); + if (!ctx) + goto out; - ctx = i915_gem_context_lookup(file->driver_priv, args->ctx_id); - if (IS_ERR(ctx)) { - mutex_unlock(&dev->struct_mutex); - return PTR_ERR(ctx); - } + /* + * We opt for unserialised reads here. This may result in tearing + * in the extremely unlikely event of a GPU hang on this context + * as we are querying them. If we need that extra layer of protection, + * we should wrap the hangstats with a seqlock. + */ if (capable(CAP_SYS_ADMIN)) args->reset_count = i915_reset_count(&dev_priv->gpu_error); else args->reset_count = 0; - args->batch_active = ctx->guilty_count; - args->batch_pending = ctx->active_count; + args->batch_active = READ_ONCE(ctx->guilty_count); + args->batch_pending = READ_ONCE(ctx->active_count); - mutex_unlock(&dev->struct_mutex); - - return 0; + ret = 0; +out: + rcu_read_unlock(); + return ret; } #if IS_ENABLED(CONFIG_DRM_I915_SELFTEST) diff --git a/drivers/gpu/drm/i915/i915_gem_context.h b/drivers/gpu/drm/i915/i915_gem_context.h index 82c99ba92ad3..04320f80f9f4 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.h +++ b/drivers/gpu/drm/i915/i915_gem_context.h @@ -86,6 +86,7 @@ struct i915_gem_context { /** link: place with &drm_i915_private.context_list */ struct list_head link; + struct llist_node free_link; /** * @ref: reference count @@ -99,6 +100,11 @@ struct i915_gem_context { struct kref ref; /** + * @rcu: rcu_head for deferred freeing. + */ + struct rcu_head rcu; + + /** * @flags: small set of booleans */ unsigned long flags; @@ -273,14 +279,18 @@ static inline bool i915_gem_context_is_kernel(struct i915_gem_context *ctx) } /* i915_gem_context.c */ -int __must_check i915_gem_context_init(struct drm_i915_private *dev_priv); -void i915_gem_context_lost(struct drm_i915_private *dev_priv); -void i915_gem_context_fini(struct drm_i915_private *dev_priv); -int i915_gem_context_open(struct drm_device *dev, struct drm_file *file); -void i915_gem_context_close(struct drm_device *dev, struct drm_file *file); +int __must_check i915_gem_contexts_init(struct drm_i915_private *dev_priv); +void i915_gem_contexts_lost(struct drm_i915_private *dev_priv); +void i915_gem_contexts_fini(struct drm_i915_private *dev_priv); + +int i915_gem_context_open(struct drm_i915_private *i915, + struct drm_file *file); +void i915_gem_context_close(struct drm_file *file); + int i915_switch_context(struct drm_i915_gem_request *req); int i915_gem_switch_to_kernel_context(struct drm_i915_private *dev_priv); -void i915_gem_context_free(struct kref *ctx_ref); + +void i915_gem_context_release(struct kref *ctx_ref); struct i915_gem_context * i915_gem_context_create_gvt(struct drm_device *dev); @@ -295,4 +305,16 @@ int i915_gem_context_setparam_ioctl(struct drm_device *dev, void *data, int i915_gem_context_reset_stats_ioctl(struct drm_device *dev, void *data, struct drm_file *file); +static inline struct i915_gem_context * +i915_gem_context_get(struct i915_gem_context *ctx) +{ + kref_get(&ctx->ref); + return ctx; +} + +static inline void i915_gem_context_put(struct i915_gem_context *ctx) +{ + kref_put(&ctx->ref, i915_gem_context_release); +} + #endif /* !__I915_GEM_CONTEXT_H__ */ diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index eb46dfa374a7..929f275e67aa 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -288,20 +288,26 @@ static int eb_create(struct i915_execbuffer *eb) * direct lookup. */ do { + unsigned int flags; + + /* While we can still reduce the allocation size, don't + * raise a warning and allow the allocation to fail. + * On the last pass though, we want to try as hard + * as possible to perform the allocation and warn + * if it fails. + */ + flags = GFP_TEMPORARY; + if (size > 1) + flags |= __GFP_NORETRY | __GFP_NOWARN; + eb->buckets = kzalloc(sizeof(struct hlist_head) << size, - GFP_TEMPORARY | - __GFP_NORETRY | - __GFP_NOWARN); + flags); if (eb->buckets) break; } while (--size); - if (unlikely(!eb->buckets)) { - eb->buckets = kzalloc(sizeof(struct hlist_head), - GFP_TEMPORARY); - if (unlikely(!eb->buckets)) - return -ENOMEM; - } + if (unlikely(!size)) + return -ENOMEM; eb->lut_size = size; } else { @@ -452,7 +458,7 @@ eb_add_vma(struct i915_execbuffer *eb, return err; } - if (eb->lut_size >= 0) { + if (eb->lut_size > 0) { vma->exec_handle = entry->handle; hlist_add_head(&vma->exec_node, &eb->buckets[hash_32(entry->handle, @@ -669,16 +675,17 @@ static int eb_select_context(struct i915_execbuffer *eb) struct i915_gem_context *ctx; ctx = i915_gem_context_lookup(eb->file->driver_priv, eb->args->rsvd1); - if (unlikely(IS_ERR(ctx))) - return PTR_ERR(ctx); + if (unlikely(!ctx)) + return -ENOENT; if (unlikely(i915_gem_context_is_banned(ctx))) { DRM_DEBUG("Context %u tried to submit while banned\n", ctx->user_handle); + i915_gem_context_put(ctx); return -EIO; } - eb->ctx = i915_gem_context_get(ctx); + eb->ctx = ctx; eb->vm = ctx->ppgtt ? &ctx->ppgtt->base : &eb->i915->ggtt.base; eb->context_flags = 0; @@ -878,6 +885,7 @@ static void eb_release_vmas(const struct i915_execbuffer *eb) GEM_BUG_ON(vma->exec_entry != entry); vma->exec_entry = NULL; + __exec_to_vma(entry) = 0; if (entry->flags & __EXEC_OBJECT_HAS_PIN) __eb_unreserve_vma(vma, entry); @@ -893,7 +901,7 @@ static void eb_release_vmas(const struct i915_execbuffer *eb) static void eb_reset_vmas(const struct i915_execbuffer *eb) { eb_release_vmas(eb); - if (eb->lut_size >= 0) + if (eb->lut_size > 0) memset(eb->buckets, 0, sizeof(struct hlist_head) << eb->lut_size); } @@ -902,7 +910,7 @@ static void eb_destroy(const struct i915_execbuffer *eb) { GEM_BUG_ON(eb->reloc_cache.rq); - if (eb->lut_size >= 0) + if (eb->lut_size > 0) kfree(eb->buckets); } @@ -1199,7 +1207,7 @@ static int __reloc_gpu_alloc(struct i915_execbuffer *eb, reservation_object_unlock(batch->resv); i915_vma_unpin(batch); - i915_vma_move_to_active(vma, rq, true); + i915_vma_move_to_active(vma, rq, EXEC_OBJECT_WRITE); reservation_object_lock(vma->resv, NULL); reservation_object_add_excl_fence(vma->resv, &rq->fence); reservation_object_unlock(vma->resv); @@ -2127,7 +2135,6 @@ i915_gem_do_execbuffer(struct drm_device *dev, if (DBG_FORCE_RELOC || !(args->flags & I915_EXEC_NO_RELOC)) args->flags |= __EXEC_HAS_RELOC; eb.exec = exec; - eb.ctx = NULL; eb.invalid_flags = __EXEC_OBJECT_UNKNOWN_FLAGS; if (USES_FULL_PPGTT(eb.i915)) eb.invalid_flags |= EXEC_OBJECT_NEEDS_GTT; @@ -2179,8 +2186,15 @@ i915_gem_do_execbuffer(struct drm_device *dev, } } - if (eb_create(&eb)) - return -ENOMEM; + err = eb_create(&eb); + if (err) + goto err_out_fence; + + GEM_BUG_ON(!eb.lut_size); + + err = eb_select_context(&eb); + if (unlikely(err)) + goto err_destroy; /* * Take a local wakeref for preparing to dispatch the execbuf as @@ -2190,14 +2204,11 @@ i915_gem_do_execbuffer(struct drm_device *dev, * 100ms. */ intel_runtime_pm_get(eb.i915); + err = i915_mutex_lock_interruptible(dev); if (err) goto err_rpm; - err = eb_select_context(&eb); - if (unlikely(err)) - goto err_unlock; - err = eb_relocate(&eb); if (err) /* @@ -2333,12 +2344,13 @@ err_batch_unpin: err_vma: if (eb.exec) eb_release_vmas(&eb); - i915_gem_context_put(eb.ctx); -err_unlock: mutex_unlock(&dev->struct_mutex); err_rpm: intel_runtime_pm_put(eb.i915); + i915_gem_context_put(eb.ctx); +err_destroy: eb_destroy(&eb); +err_out_fence: if (out_fence_fd != -1) put_unused_fd(out_fence_fd); err_in_fence: diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 61fc7e90a7da..10aa7762d9a6 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -207,8 +207,7 @@ static int ppgtt_bind_vma(struct i915_vma *vma, if (vma->obj->gt_ro) pte_flags |= PTE_READ_ONLY; - vma->vm->insert_entries(vma->vm, vma->pages, vma->node.start, - cache_level, pte_flags); + vma->vm->insert_entries(vma->vm, vma, cache_level, pte_flags); return 0; } @@ -907,37 +906,35 @@ gen8_ppgtt_insert_pte_entries(struct i915_hw_ppgtt *ppgtt, } static void gen8_ppgtt_insert_3lvl(struct i915_address_space *vm, - struct sg_table *pages, - u64 start, + struct i915_vma *vma, enum i915_cache_level cache_level, u32 unused) { struct i915_hw_ppgtt *ppgtt = i915_vm_to_ppgtt(vm); struct sgt_dma iter = { - .sg = pages->sgl, + .sg = vma->pages->sgl, .dma = sg_dma_address(iter.sg), .max = iter.dma + iter.sg->length, }; - struct gen8_insert_pte idx = gen8_insert_pte(start); + struct gen8_insert_pte idx = gen8_insert_pte(vma->node.start); gen8_ppgtt_insert_pte_entries(ppgtt, &ppgtt->pdp, &iter, &idx, cache_level); } static void gen8_ppgtt_insert_4lvl(struct i915_address_space *vm, - struct sg_table *pages, - u64 start, + struct i915_vma *vma, enum i915_cache_level cache_level, u32 unused) { struct i915_hw_ppgtt *ppgtt = i915_vm_to_ppgtt(vm); struct sgt_dma iter = { - .sg = pages->sgl, + .sg = vma->pages->sgl, .dma = sg_dma_address(iter.sg), .max = iter.dma + iter.sg->length, }; struct i915_page_directory_pointer **pdps = ppgtt->pml4.pdps; - struct gen8_insert_pte idx = gen8_insert_pte(start); + struct gen8_insert_pte idx = gen8_insert_pte(vma->node.start); while (gen8_ppgtt_insert_pte_entries(ppgtt, pdps[idx.pml4e++], &iter, &idx, cache_level)) @@ -1621,13 +1618,12 @@ static void gen6_ppgtt_clear_range(struct i915_address_space *vm, } static void gen6_ppgtt_insert_entries(struct i915_address_space *vm, - struct sg_table *pages, - u64 start, + struct i915_vma *vma, enum i915_cache_level cache_level, u32 flags) { struct i915_hw_ppgtt *ppgtt = i915_vm_to_ppgtt(vm); - unsigned first_entry = start >> PAGE_SHIFT; + unsigned first_entry = vma->node.start >> PAGE_SHIFT; unsigned act_pt = first_entry / GEN6_PTES; unsigned act_pte = first_entry % GEN6_PTES; const u32 pte_encode = vm->pte_encode(0, cache_level, flags); @@ -1635,7 +1631,7 @@ static void gen6_ppgtt_insert_entries(struct i915_address_space *vm, gen6_pte_t *vaddr; vaddr = kmap_atomic_px(ppgtt->pd.page_table[act_pt]); - iter.sg = pages->sgl; + iter.sg = vma->pages->sgl; iter.dma = sg_dma_address(iter.sg); iter.max = iter.dma + iter.sg->length; do { @@ -2090,8 +2086,7 @@ static void gen8_ggtt_insert_page(struct i915_address_space *vm, } static void gen8_ggtt_insert_entries(struct i915_address_space *vm, - struct sg_table *st, - u64 start, + struct i915_vma *vma, enum i915_cache_level level, u32 unused) { @@ -2102,8 +2097,8 @@ static void gen8_ggtt_insert_entries(struct i915_address_space *vm, dma_addr_t addr; gtt_entries = (gen8_pte_t __iomem *)ggtt->gsm; - gtt_entries += start >> PAGE_SHIFT; - for_each_sgt_dma(addr, sgt_iter, st) + gtt_entries += vma->node.start >> PAGE_SHIFT; + for_each_sgt_dma(addr, sgt_iter, vma->pages) gen8_set_pte(gtt_entries++, pte_encode | addr); wmb(); @@ -2137,17 +2132,16 @@ static void gen6_ggtt_insert_page(struct i915_address_space *vm, * mapped BAR (dev_priv->mm.gtt->gtt). */ static void gen6_ggtt_insert_entries(struct i915_address_space *vm, - struct sg_table *st, - u64 start, + struct i915_vma *vma, enum i915_cache_level level, u32 flags) { struct i915_ggtt *ggtt = i915_vm_to_ggtt(vm); gen6_pte_t __iomem *entries = (gen6_pte_t __iomem *)ggtt->gsm; - unsigned int i = start >> PAGE_SHIFT; + unsigned int i = vma->node.start >> PAGE_SHIFT; struct sgt_iter iter; dma_addr_t addr; - for_each_sgt_dma(addr, iter, st) + for_each_sgt_dma(addr, iter, vma->pages) iowrite32(vm->pte_encode(addr, level, flags), &entries[i++]); wmb(); @@ -2229,8 +2223,7 @@ static void bxt_vtd_ggtt_insert_page__BKL(struct i915_address_space *vm, struct insert_entries { struct i915_address_space *vm; - struct sg_table *st; - u64 start; + struct i915_vma *vma; enum i915_cache_level level; }; @@ -2238,19 +2231,18 @@ static int bxt_vtd_ggtt_insert_entries__cb(void *_arg) { struct insert_entries *arg = _arg; - gen8_ggtt_insert_entries(arg->vm, arg->st, arg->start, arg->level, 0); + gen8_ggtt_insert_entries(arg->vm, arg->vma, arg->level, 0); bxt_vtd_ggtt_wa(arg->vm); return 0; } static void bxt_vtd_ggtt_insert_entries__BKL(struct i915_address_space *vm, - struct sg_table *st, - u64 start, + struct i915_vma *vma, enum i915_cache_level level, u32 unused) { - struct insert_entries arg = { vm, st, start, level }; + struct insert_entries arg = { vm, vma, level }; stop_machine(bxt_vtd_ggtt_insert_entries__cb, &arg, NULL); } @@ -2316,15 +2308,15 @@ static void i915_ggtt_insert_page(struct i915_address_space *vm, } static void i915_ggtt_insert_entries(struct i915_address_space *vm, - struct sg_table *pages, - u64 start, + struct i915_vma *vma, enum i915_cache_level cache_level, u32 unused) { unsigned int flags = (cache_level == I915_CACHE_NONE) ? AGP_USER_MEMORY : AGP_USER_CACHED_MEMORY; - intel_gtt_insert_sg_entries(pages, start >> PAGE_SHIFT, flags); + intel_gtt_insert_sg_entries(vma->pages, vma->node.start >> PAGE_SHIFT, + flags); } static void i915_ggtt_clear_range(struct i915_address_space *vm, @@ -2353,8 +2345,7 @@ static int ggtt_bind_vma(struct i915_vma *vma, pte_flags |= PTE_READ_ONLY; intel_runtime_pm_get(i915); - vma->vm->insert_entries(vma->vm, vma->pages, vma->node.start, - cache_level, pte_flags); + vma->vm->insert_entries(vma->vm, vma, cache_level, pte_flags); intel_runtime_pm_put(i915); /* @@ -2407,16 +2398,13 @@ static int aliasing_gtt_bind_vma(struct i915_vma *vma, goto err_pages; } - appgtt->base.insert_entries(&appgtt->base, - vma->pages, vma->node.start, - cache_level, pte_flags); + appgtt->base.insert_entries(&appgtt->base, vma, cache_level, + pte_flags); } if (flags & I915_VMA_GLOBAL_BIND) { intel_runtime_pm_get(i915); - vma->vm->insert_entries(vma->vm, - vma->pages, vma->node.start, - cache_level, pte_flags); + vma->vm->insert_entries(vma->vm, vma, cache_level, pte_flags); intel_runtime_pm_put(i915); } diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index 1b2a56c3e5d3..b4e3aa7c0ce1 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -313,8 +313,7 @@ struct i915_address_space { enum i915_cache_level cache_level, u32 flags); void (*insert_entries)(struct i915_address_space *vm, - struct sg_table *st, - u64 start, + struct i915_vma *vma, enum i915_cache_level cache_level, u32 flags); void (*cleanup)(struct i915_address_space *vm); diff --git a/drivers/gpu/drm/i915/i915_gem_request.c b/drivers/gpu/drm/i915/i915_gem_request.c index 8c59c79cbd8b..483af8921060 100644 --- a/drivers/gpu/drm/i915/i915_gem_request.c +++ b/drivers/gpu/drm/i915/i915_gem_request.c @@ -384,7 +384,11 @@ static void i915_gem_request_retire(struct drm_i915_gem_request *request) engine->context_unpin(engine, engine->last_retired_context); engine->last_retired_context = request->ctx; - dma_fence_signal(&request->fence); + spin_lock_irq(&request->lock); + if (request->waitboost) + atomic_dec(&request->i915->rps.num_waiters); + dma_fence_signal_locked(&request->fence); + spin_unlock_irq(&request->lock); i915_priotree_fini(request->i915, &request->priotree); i915_gem_request_put(request); @@ -639,6 +643,7 @@ i915_gem_request_alloc(struct intel_engine_cs *engine, req->file_priv = NULL; req->batch = NULL; req->capture_list = NULL; + req->waitboost = false; /* * Reserve space in the ring buffer for all the commands required to diff --git a/drivers/gpu/drm/i915/i915_gem_request.h b/drivers/gpu/drm/i915/i915_gem_request.h index 7b7c84369d78..49a4c8994ff0 100644 --- a/drivers/gpu/drm/i915/i915_gem_request.h +++ b/drivers/gpu/drm/i915/i915_gem_request.h @@ -129,7 +129,7 @@ struct drm_i915_gem_request { * It is used by the driver to then queue the request for execution. */ struct i915_sw_fence submit; - wait_queue_t submitq; + wait_queue_entry_t submitq; wait_queue_head_t execute; /* A list of everyone we wait upon, and everyone who waits upon us. @@ -184,6 +184,8 @@ struct drm_i915_gem_request { /** Time at which this request was emitted, in jiffies. */ unsigned long emitted_jiffies; + bool waitboost; + /** engine->request_list entry for this request */ struct list_head link; diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index e18f350bc364..ae70283470a6 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -463,6 +463,7 @@ static void error_print_engine(struct drm_i915_error_state_buf *m, err_printf(m, " hangcheck action timestamp: %lu, %u ms ago\n", ee->hangcheck_timestamp, jiffies_to_msecs(jiffies - ee->hangcheck_timestamp)); + err_printf(m, " engine reset count: %u\n", ee->reset_count); error_print_request(m, " ELSP[0]: ", &ee->execlist[0]); error_print_request(m, " ELSP[1]: ", &ee->execlist[1]); @@ -1236,6 +1237,8 @@ static void error_record_engine_registers(struct i915_gpu_state *error, ee->hangcheck_timestamp = engine->hangcheck.action_timestamp; ee->hangcheck_action = engine->hangcheck.action; ee->hangcheck_stalled = engine->hangcheck.stalled; + ee->reset_count = i915_reset_engine_count(&dev_priv->gpu_error, + engine); if (USES_PPGTT(dev_priv)) { int i; diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index bce2d1feceb1..eb4f1dca2077 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -1091,18 +1091,6 @@ static u32 vlv_wa_c0_ei(struct drm_i915_private *dev_priv, u32 pm_iir) return events; } -static bool any_waiters(struct drm_i915_private *dev_priv) -{ - struct intel_engine_cs *engine; - enum intel_engine_id id; - - for_each_engine(engine, dev_priv, id) - if (intel_engine_has_waiter(engine)) - return true; - - return false; -} - static void gen6_pm_rps_work(struct work_struct *work) { struct drm_i915_private *dev_priv = @@ -1114,7 +1102,7 @@ static void gen6_pm_rps_work(struct work_struct *work) spin_lock_irq(&dev_priv->irq_lock); if (dev_priv->rps.interrupts_enabled) { pm_iir = fetch_and_zero(&dev_priv->rps.pm_iir); - client_boost = fetch_and_zero(&dev_priv->rps.client_boost); + client_boost = atomic_read(&dev_priv->rps.num_waiters); } spin_unlock_irq(&dev_priv->irq_lock); @@ -1131,7 +1119,7 @@ static void gen6_pm_rps_work(struct work_struct *work) new_delay = dev_priv->rps.cur_freq; min = dev_priv->rps.min_freq_softlimit; max = dev_priv->rps.max_freq_softlimit; - if (client_boost || any_waiters(dev_priv)) + if (client_boost) max = dev_priv->rps.max_freq; if (client_boost && new_delay < dev_priv->rps.boost_freq) { new_delay = dev_priv->rps.boost_freq; @@ -1144,7 +1132,7 @@ static void gen6_pm_rps_work(struct work_struct *work) if (new_delay >= dev_priv->rps.max_freq_softlimit) adj = 0; - } else if (client_boost || any_waiters(dev_priv)) { + } else if (client_boost) { adj = 0; } else if (pm_iir & GEN6_PM_RP_DOWN_TIMEOUT) { if (dev_priv->rps.cur_freq > dev_priv->rps.efficient_freq) @@ -2599,60 +2587,93 @@ static irqreturn_t gen8_irq_handler(int irq, void *arg) return ret; } +struct wedge_me { + struct delayed_work work; + struct drm_i915_private *i915; + const char *name; +}; + +static void wedge_me(struct work_struct *work) +{ + struct wedge_me *w = container_of(work, typeof(*w), work.work); + + dev_err(w->i915->drm.dev, + "%s timed out, cancelling all in-flight rendering.\n", + w->name); + i915_gem_set_wedged(w->i915); +} + +static void __init_wedge(struct wedge_me *w, + struct drm_i915_private *i915, + long timeout, + const char *name) +{ + w->i915 = i915; + w->name = name; + + INIT_DELAYED_WORK_ONSTACK(&w->work, wedge_me); + schedule_delayed_work(&w->work, timeout); +} + +static void __fini_wedge(struct wedge_me *w) +{ + cancel_delayed_work_sync(&w->work); + destroy_delayed_work_on_stack(&w->work); + w->i915 = NULL; +} + +#define i915_wedge_on_timeout(W, DEV, TIMEOUT) \ + for (__init_wedge((W), (DEV), (TIMEOUT), __func__); \ + (W)->i915; \ + __fini_wedge((W))) + /** - * i915_reset_and_wakeup - do process context error handling work + * i915_reset_device - do process context error handling work * @dev_priv: i915 device private * * Fire an error uevent so userspace can see that a hang or error * was detected. */ -static void i915_reset_and_wakeup(struct drm_i915_private *dev_priv) +static void i915_reset_device(struct drm_i915_private *dev_priv) { struct kobject *kobj = &dev_priv->drm.primary->kdev->kobj; char *error_event[] = { I915_ERROR_UEVENT "=1", NULL }; char *reset_event[] = { I915_RESET_UEVENT "=1", NULL }; char *reset_done_event[] = { I915_ERROR_UEVENT "=0", NULL }; + struct wedge_me w; kobject_uevent_env(kobj, KOBJ_CHANGE, error_event); DRM_DEBUG_DRIVER("resetting chip\n"); kobject_uevent_env(kobj, KOBJ_CHANGE, reset_event); - intel_prepare_reset(dev_priv); + /* Use a watchdog to ensure that our reset completes */ + i915_wedge_on_timeout(&w, dev_priv, 5*HZ) { + intel_prepare_reset(dev_priv); - set_bit(I915_RESET_HANDOFF, &dev_priv->gpu_error.flags); - wake_up_all(&dev_priv->gpu_error.wait_queue); + /* Signal that locked waiters should reset the GPU */ + set_bit(I915_RESET_HANDOFF, &dev_priv->gpu_error.flags); + wake_up_all(&dev_priv->gpu_error.wait_queue); - do { - /* - * All state reset _must_ be completed before we update the - * reset counter, for otherwise waiters might miss the reset - * pending state and not properly drop locks, resulting in - * deadlocks with the reset work. + /* Wait for anyone holding the lock to wakeup, without + * blocking indefinitely on struct_mutex. */ - if (mutex_trylock(&dev_priv->drm.struct_mutex)) { - i915_reset(dev_priv); - mutex_unlock(&dev_priv->drm.struct_mutex); - } - - /* We need to wait for anyone holding the lock to wakeup */ - } while (wait_on_bit_timeout(&dev_priv->gpu_error.flags, - I915_RESET_HANDOFF, - TASK_UNINTERRUPTIBLE, - HZ)); + do { + if (mutex_trylock(&dev_priv->drm.struct_mutex)) { + i915_reset(dev_priv); + mutex_unlock(&dev_priv->drm.struct_mutex); + } + } while (wait_on_bit_timeout(&dev_priv->gpu_error.flags, + I915_RESET_HANDOFF, + TASK_UNINTERRUPTIBLE, + 1)); - intel_finish_reset(dev_priv); + intel_finish_reset(dev_priv); + } if (!test_bit(I915_WEDGED, &dev_priv->gpu_error.flags)) kobject_uevent_env(kobj, KOBJ_CHANGE, reset_done_event); - - /* - * Note: The wake_up also serves as a memory barrier so that - * waiters see the updated value of the dev_priv->gpu_error. - */ - clear_bit(I915_RESET_BACKOFF, &dev_priv->gpu_error.flags); - wake_up_all(&dev_priv->gpu_error.reset_queue); } static inline void @@ -2722,6 +2743,8 @@ void i915_handle_error(struct drm_i915_private *dev_priv, u32 engine_mask, const char *fmt, ...) { + struct intel_engine_cs *engine; + unsigned int tmp; va_list args; char error_msg[80]; @@ -2741,14 +2764,56 @@ void i915_handle_error(struct drm_i915_private *dev_priv, i915_capture_error_state(dev_priv, engine_mask, error_msg); i915_clear_error_registers(dev_priv); + /* + * Try engine reset when available. We fall back to full reset if + * single reset fails. + */ + if (intel_has_reset_engine(dev_priv)) { + for_each_engine_masked(engine, dev_priv, engine_mask, tmp) { + BUILD_BUG_ON(I915_RESET_HANDOFF >= I915_RESET_ENGINE); + if (test_and_set_bit(I915_RESET_ENGINE + engine->id, + &dev_priv->gpu_error.flags)) + continue; + + if (i915_reset_engine(engine) == 0) + engine_mask &= ~intel_engine_flag(engine); + + clear_bit(I915_RESET_ENGINE + engine->id, + &dev_priv->gpu_error.flags); + wake_up_bit(&dev_priv->gpu_error.flags, + I915_RESET_ENGINE + engine->id); + } + } + if (!engine_mask) goto out; - if (test_and_set_bit(I915_RESET_BACKOFF, - &dev_priv->gpu_error.flags)) + /* Full reset needs the mutex, stop any other user trying to do so. */ + if (test_and_set_bit(I915_RESET_BACKOFF, &dev_priv->gpu_error.flags)) { + wait_event(dev_priv->gpu_error.reset_queue, + !test_bit(I915_RESET_BACKOFF, + &dev_priv->gpu_error.flags)); goto out; + } + + /* Prevent any other reset-engine attempt. */ + for_each_engine(engine, dev_priv, tmp) { + while (test_and_set_bit(I915_RESET_ENGINE + engine->id, + &dev_priv->gpu_error.flags)) + wait_on_bit(&dev_priv->gpu_error.flags, + I915_RESET_ENGINE + engine->id, + TASK_UNINTERRUPTIBLE); + } - i915_reset_and_wakeup(dev_priv); + i915_reset_device(dev_priv); + + for_each_engine(engine, dev_priv, tmp) { + clear_bit(I915_RESET_ENGINE + engine->id, + &dev_priv->gpu_error.flags); + } + + clear_bit(I915_RESET_BACKOFF, &dev_priv->gpu_error.flags); + wake_up_all(&dev_priv->gpu_error.reset_queue); out: intel_runtime_pm_put(dev_priv); diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c index b6a7e363d076..88b9d3e6713a 100644 --- a/drivers/gpu/drm/i915/i915_params.c +++ b/drivers/gpu/drm/i915/i915_params.c @@ -46,7 +46,7 @@ struct i915_params i915 __read_mostly = { .prefault_disable = 0, .load_detect_test = 0, .force_reset_modeset_test = 0, - .reset = true, + .reset = 2, .error_capture = true, .invert_brightness = 0, .disable_display = 0, @@ -63,8 +63,9 @@ struct i915_params i915 __read_mostly = { .huc_firmware_path = NULL, .enable_dp_mst = true, .inject_load_failure = 0, - .enable_dpcd_backlight = false, + .enable_dpcd_backlight = -1, .enable_gvt = false, + .enable_dbc = true, }; module_param_named(modeset, i915.modeset, int, 0400); @@ -115,8 +116,8 @@ MODULE_PARM_DESC(vbt_sdvo_panel_type, "Override/Ignore selection of SDVO panel mode in the VBT " "(-2=ignore, -1=auto [default], index in VBT BIOS table)"); -module_param_named_unsafe(reset, i915.reset, bool, 0600); -MODULE_PARM_DESC(reset, "Attempt GPU resets (default: true)"); +module_param_named_unsafe(reset, i915.reset, int, 0600); +MODULE_PARM_DESC(reset, "Attempt GPU resets (0=disabled, 1=full gpu reset, 2=engine reset [default])"); #if IS_ENABLED(CONFIG_DRM_I915_CAPTURE_ERROR) module_param_named(error_capture, i915.error_capture, bool, 0600); @@ -246,10 +247,15 @@ MODULE_PARM_DESC(enable_dp_mst, module_param_named_unsafe(inject_load_failure, i915.inject_load_failure, uint, 0400); MODULE_PARM_DESC(inject_load_failure, "Force an error after a number of failure check points (0:disabled (default), N:force failure at the Nth failure check point)"); -module_param_named(enable_dpcd_backlight, i915.enable_dpcd_backlight, bool, 0600); +module_param_named_unsafe(enable_dpcd_backlight, i915.enable_dpcd_backlight, int, 0600); MODULE_PARM_DESC(enable_dpcd_backlight, - "Enable support for DPCD backlight control (default:false)"); + "Enable support for DPCD backlight control " + "(-1:auto (default), 0:force disable, 1:force enabled if supported"); module_param_named(enable_gvt, i915.enable_gvt, bool, 0400); MODULE_PARM_DESC(enable_gvt, "Enable support for Intel GVT-g graphics virtualization host support(default:false)"); + +module_param_named_unsafe(enable_dbc, i915.enable_dbc, bool, 0600); +MODULE_PARM_DESC(enable_dbc, + "Enable support for dynamic backlight control (default:true)"); diff --git a/drivers/gpu/drm/i915/i915_params.h b/drivers/gpu/drm/i915/i915_params.h index 34148cc8637c..057e203e6bda 100644 --- a/drivers/gpu/drm/i915/i915_params.h +++ b/drivers/gpu/drm/i915/i915_params.h @@ -51,7 +51,9 @@ func(int, use_mmio_flip); \ func(int, mmio_debug); \ func(int, edp_vswing); \ + func(int, reset); \ func(unsigned int, inject_load_failure); \ + func(int, enable_dpcd_backlight); \ /* leave bools at the end to not create holes */ \ func(bool, alpha_support); \ func(bool, enable_cmd_parser); \ @@ -60,14 +62,13 @@ func(bool, prefault_disable); \ func(bool, load_detect_test); \ func(bool, force_reset_modeset_test); \ - func(bool, reset); \ func(bool, error_capture); \ func(bool, disable_display); \ func(bool, verbose_state_checks); \ func(bool, nuclear_pageflip); \ func(bool, enable_dp_mst); \ - func(bool, enable_dpcd_backlight); \ - func(bool, enable_gvt) + func(bool, enable_gvt); \ + func(bool, enable_dbc) #define MEMBER(T, member) T member struct i915_params { diff --git a/drivers/gpu/drm/i915/i915_pci.c b/drivers/gpu/drm/i915/i915_pci.c index 506ec32b9e53..a1e6b696bcfa 100644 --- a/drivers/gpu/drm/i915/i915_pci.c +++ b/drivers/gpu/drm/i915/i915_pci.c @@ -310,7 +310,8 @@ static const struct intel_device_info intel_haswell_info = { BDW_COLORS, \ .has_logical_ring_contexts = 1, \ .has_full_48bit_ppgtt = 1, \ - .has_64bit_reloc = 1 + .has_64bit_reloc = 1, \ + .has_reset_engine = 1 #define BDW_PLATFORM \ BDW_FEATURES, \ @@ -342,6 +343,7 @@ static const struct intel_device_info intel_cherryview_info = { .has_gmch_display = 1, .has_aliasing_ppgtt = 1, .has_full_ppgtt = 1, + .has_reset_engine = 1, .display_mmio_offset = VLV_DISPLAY_BASE, GEN_CHV_PIPEOFFSETS, CURSOR_OFFSETS, @@ -387,6 +389,7 @@ static const struct intel_device_info intel_skylake_gt3_info = { .has_aliasing_ppgtt = 1, \ .has_full_ppgtt = 1, \ .has_full_48bit_ppgtt = 1, \ + .has_reset_engine = 1, \ GEN_DEFAULT_PIPEOFFSETS, \ IVB_CURSOR_OFFSETS, \ BDW_COLORS @@ -446,6 +449,7 @@ static const struct intel_device_info intel_cannonlake_info = { .gen = 10, .ddb_size = 1024, .has_csr = 1, + .color = { .degamma_lut_size = 0, .gamma_lut_size = 1024 } }; /* diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c index 38c44407bafc..d9f77a4d85db 100644 --- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -1746,7 +1746,7 @@ static int gen8_configure_all_contexts(struct drm_i915_private *dev_priv, goto out; /* Update all contexts now that we've stalled the submission. */ - list_for_each_entry(ctx, &dev_priv->context_list, link) { + list_for_each_entry(ctx, &dev_priv->contexts.list, link) { struct intel_context *ce = &ctx->engine[RCS]; u32 *regs; @@ -2067,10 +2067,6 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, return ret; } - ret = alloc_oa_buffer(dev_priv); - if (ret) - goto err_oa_buf_alloc; - /* PRM - observability performance counters: * * OACONTROL, performance counter enable, note: @@ -2086,6 +2082,10 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, intel_runtime_pm_get(dev_priv); intel_uncore_forcewake_get(dev_priv, FORCEWAKE_ALL); + ret = alloc_oa_buffer(dev_priv); + if (ret) + goto err_oa_buf_alloc; + ret = dev_priv->perf.oa.ops.enable_metric_set(dev_priv); if (ret) goto err_enable; @@ -2097,11 +2097,11 @@ static int i915_oa_stream_init(struct i915_perf_stream *stream, return 0; err_enable: - intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); - intel_runtime_pm_put(dev_priv); free_oa_buffer(dev_priv); err_oa_buf_alloc: + intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); + intel_runtime_pm_put(dev_priv); if (stream->ctx) oa_put_render_ctx_id(stream); @@ -2444,7 +2444,7 @@ static void i915_perf_destroy_locked(struct i915_perf_stream *stream) list_del(&stream->link); if (stream->ctx) - i915_gem_context_put_unlocked(stream->ctx); + i915_gem_context_put(stream->ctx); kfree(stream); } @@ -2633,7 +2633,7 @@ err_alloc: kfree(stream); err_ctx: if (specific_ctx) - i915_gem_context_put_unlocked(specific_ctx); + i915_gem_context_put(specific_ctx); err: return ret; } diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index bd535f12db18..c712d01f92ab 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -1764,8 +1764,11 @@ enum skl_disp_power_wells { _CNL_PORT_TX_DW2_LN0_AE, \ _CNL_PORT_TX_DW2_LN0_F) #define SWING_SEL_UPPER(x) ((x >> 3) << 15) +#define SWING_SEL_UPPER_MASK (1 << 15) #define SWING_SEL_LOWER(x) ((x & 0x7) << 11) +#define SWING_SEL_LOWER_MASK (0x7 << 11) #define RCOMP_SCALAR(x) ((x) << 0) +#define RCOMP_SCALAR_MASK (0xFF << 0) #define _CNL_PORT_TX_DW4_GRP_AE 0x162350 #define _CNL_PORT_TX_DW4_GRP_B 0x1623D0 @@ -1795,8 +1798,11 @@ enum skl_disp_power_wells { _CNL_PORT_TX_DW4_LN0_F) #define LOADGEN_SELECT (1 << 31) #define POST_CURSOR_1(x) ((x) << 12) +#define POST_CURSOR_1_MASK (0x3F << 12) #define POST_CURSOR_2(x) ((x) << 6) +#define POST_CURSOR_2_MASK (0x3F << 6) #define CURSOR_COEFF(x) ((x) << 0) +#define CURSOR_COEFF_MASK (0x3F << 0) #define _CNL_PORT_TX_DW5_GRP_AE 0x162354 #define _CNL_PORT_TX_DW5_GRP_B 0x1623D4 @@ -1825,7 +1831,9 @@ enum skl_disp_power_wells { #define TX_TRAINING_EN (1 << 31) #define TAP3_DISABLE (1 << 29) #define SCALING_MODE_SEL(x) ((x) << 18) +#define SCALING_MODE_SEL_MASK (0x7 << 18) #define RTERM_SELECT(x) ((x) << 3) +#define RTERM_SELECT_MASK (0x7 << 3) #define _CNL_PORT_TX_DW7_GRP_AE 0x16235C #define _CNL_PORT_TX_DW7_GRP_B 0x1623DC @@ -1852,6 +1860,7 @@ enum skl_disp_power_wells { _CNL_PORT_TX_DW7_LN0_AE, \ _CNL_PORT_TX_DW7_LN0_F) #define N_SCALAR(x) ((x) << 24) +#define N_SCALAR_MASK (0x7F << 24) /* The spec defines this only for BXT PHY0, but lets assume that this * would exist for PHY1 too if it had a second channel. @@ -3513,7 +3522,7 @@ enum skl_disp_power_wells { #define INTERVAL_1_28_US(us) roundup(((us) * 100) >> 7, 25) #define INTERVAL_1_33_US(us) (((us) * 3) >> 2) #define INTERVAL_0_833_US(us) (((us) * 6) / 5) -#define GT_INTERVAL_FROM_US(dev_priv, us) (IS_GEN9(dev_priv) ? \ +#define GT_INTERVAL_FROM_US(dev_priv, us) (INTEL_GEN(dev_priv) >= 9 ? \ (IS_GEN9_LP(dev_priv) ? \ INTERVAL_0_833_US(us) : \ INTERVAL_1_33_US(us)) : \ @@ -3522,7 +3531,7 @@ enum skl_disp_power_wells { #define INTERVAL_1_28_TO_US(interval) (((interval) << 7) / 100) #define INTERVAL_1_33_TO_US(interval) (((interval) << 2) / 3) #define INTERVAL_0_833_TO_US(interval) (((interval) * 5) / 6) -#define GT_PM_INTERVAL_TO_US(dev_priv, interval) (IS_GEN9(dev_priv) ? \ +#define GT_PM_INTERVAL_TO_US(dev_priv, interval) (INTEL_GEN(dev_priv) >= 9 ? \ (IS_GEN9_LP(dev_priv) ? \ INTERVAL_0_833_TO_US(interval) : \ INTERVAL_1_33_TO_US(interval)) : \ @@ -8334,6 +8343,7 @@ enum { #define DPLL_CFGCR0_LINK_RATE_3240 (6 << 25) #define DPLL_CFGCR0_LINK_RATE_4050 (7 << 25) #define DPLL_CFGCR0_DCO_FRACTION_MASK (0x7fff << 10) +#define DPLL_CFGCR0_DCO_FRAC_SHIFT (10) #define DPLL_CFGCR0_DCO_FRACTION(x) ((x) << 10) #define DPLL_CFGCR0_DCO_INTEGER_MASK (0x3ff) #define CNL_DPLL_CFGCR0(pll) _MMIO_PLL(pll, _CNL_DPLL0_CFGCR0, _CNL_DPLL1_CFGCR0) @@ -8341,6 +8351,7 @@ enum { #define _CNL_DPLL0_CFGCR1 0x6C004 #define _CNL_DPLL1_CFGCR1 0x6C084 #define DPLL_CFGCR1_QDIV_RATIO_MASK (0xff << 10) +#define DPLL_CFGCR1_QDIV_RATIO_SHIFT (10) #define DPLL_CFGCR1_QDIV_RATIO(x) ((x) << 10) #define DPLL_CFGCR1_QDIV_MODE(x) ((x) << 9) #define DPLL_CFGCR1_KDIV_MASK (7 << 6) diff --git a/drivers/gpu/drm/i915/i915_sw_fence.c b/drivers/gpu/drm/i915/i915_sw_fence.c index 474d23c0c0ce..f29540f922af 100644 --- a/drivers/gpu/drm/i915/i915_sw_fence.c +++ b/drivers/gpu/drm/i915/i915_sw_fence.c @@ -125,7 +125,7 @@ static void __i915_sw_fence_wake_up_all(struct i915_sw_fence *fence, struct list_head *continuation) { wait_queue_head_t *x = &fence->wait; - wait_queue_t *pos, *next; + wait_queue_entry_t *pos, *next; unsigned long flags; debug_fence_deactivate(fence); @@ -133,31 +133,30 @@ static void __i915_sw_fence_wake_up_all(struct i915_sw_fence *fence, /* * To prevent unbounded recursion as we traverse the graph of - * i915_sw_fences, we move the task_list from this, the next ready - * fence, to the tail of the original fence's task_list + * i915_sw_fences, we move the entry list from this, the next ready + * fence, to the tail of the original fence's entry list * (and so added to the list to be woken). */ spin_lock_irqsave_nested(&x->lock, flags, 1 + !!continuation); if (continuation) { - list_for_each_entry_safe(pos, next, &x->task_list, task_list) { + list_for_each_entry_safe(pos, next, &x->head, entry) { if (pos->func == autoremove_wake_function) pos->func(pos, TASK_NORMAL, 0, continuation); else - list_move_tail(&pos->task_list, continuation); + list_move_tail(&pos->entry, continuation); } } else { LIST_HEAD(extra); do { - list_for_each_entry_safe(pos, next, - &x->task_list, task_list) + list_for_each_entry_safe(pos, next, &x->head, entry) pos->func(pos, TASK_NORMAL, 0, &extra); if (list_empty(&extra)) break; - list_splice_tail_init(&extra, &x->task_list); + list_splice_tail_init(&extra, &x->head); } while (1); } spin_unlock_irqrestore(&x->lock, flags); @@ -222,9 +221,9 @@ void i915_sw_fence_commit(struct i915_sw_fence *fence) i915_sw_fence_complete(fence); } -static int i915_sw_fence_wake(wait_queue_t *wq, unsigned mode, int flags, void *key) +static int i915_sw_fence_wake(wait_queue_entry_t *wq, unsigned mode, int flags, void *key) { - list_del(&wq->task_list); + list_del(&wq->entry); __i915_sw_fence_complete(wq->private, key); if (wq->flags & I915_SW_FENCE_FLAG_ALLOC) @@ -235,7 +234,7 @@ static int i915_sw_fence_wake(wait_queue_t *wq, unsigned mode, int flags, void * static bool __i915_sw_fence_check_if_after(struct i915_sw_fence *fence, const struct i915_sw_fence * const signaler) { - wait_queue_t *wq; + wait_queue_entry_t *wq; if (__test_and_set_bit(I915_SW_FENCE_CHECKED_BIT, &fence->flags)) return false; @@ -243,7 +242,7 @@ static bool __i915_sw_fence_check_if_after(struct i915_sw_fence *fence, if (fence == signaler) return true; - list_for_each_entry(wq, &fence->wait.task_list, task_list) { + list_for_each_entry(wq, &fence->wait.head, entry) { if (wq->func != i915_sw_fence_wake) continue; @@ -256,12 +255,12 @@ static bool __i915_sw_fence_check_if_after(struct i915_sw_fence *fence, static void __i915_sw_fence_clear_checked_bit(struct i915_sw_fence *fence) { - wait_queue_t *wq; + wait_queue_entry_t *wq; if (!__test_and_clear_bit(I915_SW_FENCE_CHECKED_BIT, &fence->flags)) return; - list_for_each_entry(wq, &fence->wait.task_list, task_list) { + list_for_each_entry(wq, &fence->wait.head, entry) { if (wq->func != i915_sw_fence_wake) continue; @@ -288,7 +287,7 @@ static bool i915_sw_fence_check_if_after(struct i915_sw_fence *fence, static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence, struct i915_sw_fence *signaler, - wait_queue_t *wq, gfp_t gfp) + wait_queue_entry_t *wq, gfp_t gfp) { unsigned long flags; int pending; @@ -318,7 +317,7 @@ static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence, pending |= I915_SW_FENCE_FLAG_ALLOC; } - INIT_LIST_HEAD(&wq->task_list); + INIT_LIST_HEAD(&wq->entry); wq->flags = pending; wq->func = i915_sw_fence_wake; wq->private = fence; @@ -327,7 +326,7 @@ static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence, spin_lock_irqsave(&signaler->wait.lock, flags); if (likely(!i915_sw_fence_done(signaler))) { - __add_wait_queue_tail(&signaler->wait, wq); + __add_wait_queue_entry_tail(&signaler->wait, wq); pending = 1; } else { i915_sw_fence_wake(wq, 0, 0, NULL); @@ -340,7 +339,7 @@ static int __i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence, int i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence, struct i915_sw_fence *signaler, - wait_queue_t *wq) + wait_queue_entry_t *wq) { return __i915_sw_fence_await_sw_fence(fence, signaler, wq, 0); } diff --git a/drivers/gpu/drm/i915/i915_sw_fence.h b/drivers/gpu/drm/i915/i915_sw_fence.h index 1d3b6051daaf..fe2ef4dadfc6 100644 --- a/drivers/gpu/drm/i915/i915_sw_fence.h +++ b/drivers/gpu/drm/i915/i915_sw_fence.h @@ -65,7 +65,7 @@ void i915_sw_fence_commit(struct i915_sw_fence *fence); int i915_sw_fence_await_sw_fence(struct i915_sw_fence *fence, struct i915_sw_fence *after, - wait_queue_t *wq); + wait_queue_entry_t *wq); int i915_sw_fence_await_sw_fence_gfp(struct i915_sw_fence *fence, struct i915_sw_fence *after, gfp_t gfp); diff --git a/drivers/gpu/drm/i915/i915_sysfs.c b/drivers/gpu/drm/i915/i915_sysfs.c index 1eef3fae4db3..7fcf00622c4c 100644 --- a/drivers/gpu/drm/i915/i915_sysfs.c +++ b/drivers/gpu/drm/i915/i915_sysfs.c @@ -96,7 +96,7 @@ static struct attribute *rc6_attrs[] = { NULL }; -static struct attribute_group rc6_attr_group = { +static const struct attribute_group rc6_attr_group = { .name = power_group_name, .attrs = rc6_attrs }; @@ -107,7 +107,7 @@ static struct attribute *rc6p_attrs[] = { NULL }; -static struct attribute_group rc6p_attr_group = { +static const struct attribute_group rc6p_attr_group = { .name = power_group_name, .attrs = rc6p_attrs }; @@ -117,7 +117,7 @@ static struct attribute *media_rc6_attrs[] = { NULL }; -static struct attribute_group media_rc6_attr_group = { +static const struct attribute_group media_rc6_attr_group = { .name = power_group_name, .attrs = media_rc6_attrs }; @@ -209,7 +209,7 @@ i915_l3_write(struct file *filp, struct kobject *kobj, memcpy(*remap_info + (offset/4), buf, count); /* NB: We defer the remapping until we switch to the context */ - list_for_each_entry(ctx, &dev_priv->context_list, link) + list_for_each_entry(ctx, &dev_priv->contexts.list, link) ctx->remap_slice |= (1<<slice); ret = count; @@ -253,7 +253,7 @@ static ssize_t gt_act_freq_mhz_show(struct device *kdev, ret = intel_gpu_freq(dev_priv, (freq >> 8) & 0xff); } else { u32 rpstat = I915_READ(GEN6_RPSTAT1); - if (IS_GEN9(dev_priv)) + if (INTEL_GEN(dev_priv) >= 9) ret = (rpstat & GEN9_CAGF_MASK) >> GEN9_CAGF_SHIFT; else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) ret = (rpstat & HSW_CAGF_MASK) >> HSW_CAGF_SHIFT; diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c index 532c709febbd..958be0a95960 100644 --- a/drivers/gpu/drm/i915/i915_vma.c +++ b/drivers/gpu/drm/i915/i915_vma.c @@ -579,11 +579,17 @@ err_unpin: static void i915_vma_destroy(struct i915_vma *vma) { + int i; + GEM_BUG_ON(vma->node.allocated); GEM_BUG_ON(i915_vma_is_active(vma)); GEM_BUG_ON(!i915_vma_is_closed(vma)); GEM_BUG_ON(vma->fence); + for (i = 0; i < ARRAY_SIZE(vma->last_read); i++) + GEM_BUG_ON(i915_gem_active_isset(&vma->last_read[i])); + GEM_BUG_ON(i915_gem_active_isset(&vma->last_fence)); + list_del(&vma->vm_link); if (!i915_vma_is_ggtt(vma)) i915_ppgtt_put(i915_vm_to_ppgtt(vma->vm)); @@ -672,12 +678,16 @@ int i915_vma_unbind(struct i915_vma *vma) break; } + if (!ret) { + ret = i915_gem_active_retire(&vma->last_fence, + &vma->vm->i915->drm.struct_mutex); + } + __i915_vma_unpin(vma); if (ret) return ret; - - GEM_BUG_ON(i915_vma_is_active(vma)); } + GEM_BUG_ON(i915_vma_is_active(vma)); if (i915_vma_is_pinned(vma)) return -EBUSY; diff --git a/drivers/gpu/drm/i915/intel_acpi.c b/drivers/gpu/drm/i915/intel_acpi.c index eb638a1e69d2..42fb436f6cdc 100644 --- a/drivers/gpu/drm/i915/intel_acpi.c +++ b/drivers/gpu/drm/i915/intel_acpi.c @@ -15,13 +15,9 @@ static struct intel_dsm_priv { acpi_handle dhandle; } intel_dsm_priv; -static const u8 intel_dsm_guid[] = { - 0xd3, 0x73, 0xd8, 0x7e, - 0xd0, 0xc2, - 0x4f, 0x4e, - 0xa8, 0x54, - 0x0f, 0x13, 0x17, 0xb0, 0x1c, 0x2c -}; +static const guid_t intel_dsm_guid = + GUID_INIT(0x7ed873d3, 0xc2d0, 0x4e4f, + 0xa8, 0x54, 0x0f, 0x13, 0x17, 0xb0, 0x1c, 0x2c); static char *intel_dsm_port_name(u8 id) { @@ -80,7 +76,7 @@ static void intel_dsm_platform_mux_info(void) int i; union acpi_object *pkg, *connector_count; - pkg = acpi_evaluate_dsm_typed(intel_dsm_priv.dhandle, intel_dsm_guid, + pkg = acpi_evaluate_dsm_typed(intel_dsm_priv.dhandle, &intel_dsm_guid, INTEL_DSM_REVISION_ID, INTEL_DSM_FN_PLATFORM_MUX_INFO, NULL, ACPI_TYPE_PACKAGE); if (!pkg) { @@ -118,7 +114,7 @@ static bool intel_dsm_pci_probe(struct pci_dev *pdev) if (!dhandle) return false; - if (!acpi_check_dsm(dhandle, intel_dsm_guid, INTEL_DSM_REVISION_ID, + if (!acpi_check_dsm(dhandle, &intel_dsm_guid, INTEL_DSM_REVISION_ID, 1 << INTEL_DSM_FN_PLATFORM_MUX_INFO)) { DRM_DEBUG_KMS("no _DSM method for intel device\n"); return false; diff --git a/drivers/gpu/drm/i915/intel_atomic_plane.c b/drivers/gpu/drm/i915/intel_atomic_plane.c index 4325cb0a04f5..ee76fab7bb6f 100644 --- a/drivers/gpu/drm/i915/intel_atomic_plane.c +++ b/drivers/gpu/drm/i915/intel_atomic_plane.c @@ -114,6 +114,8 @@ int intel_plane_atomic_check_with_state(struct intel_crtc_state *crtc_state, struct drm_i915_private *dev_priv = to_i915(plane->dev); struct drm_plane_state *state = &intel_state->base; struct intel_plane *intel_plane = to_intel_plane(plane); + const struct drm_display_mode *adjusted_mode = + &crtc_state->base.adjusted_mode; int ret; /* @@ -173,6 +175,19 @@ int intel_plane_atomic_check_with_state(struct intel_crtc_state *crtc_state, if (ret) return ret; + /* + * Y-tiling is not supported in IF-ID Interlace mode in + * GEN9 and above. + */ + if (state->fb && INTEL_GEN(dev_priv) >= 9 && crtc_state->base.enable && + adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE) { + if (state->fb->modifier == I915_FORMAT_MOD_Y_TILED || + state->fb->modifier == I915_FORMAT_MOD_Yf_TILED) { + DRM_DEBUG_KMS("Y/Yf tiling not supported in IF-ID mode\n"); + return -EINVAL; + } + } + /* FIXME pre-g4x don't work like this */ if (intel_state->base.visible) crtc_state->active_planes |= BIT(intel_plane->id); diff --git a/drivers/gpu/drm/i915/intel_bios.c b/drivers/gpu/drm/i915/intel_bios.c index 639d45c1dd2e..82b144cdfa1d 100644 --- a/drivers/gpu/drm/i915/intel_bios.c +++ b/drivers/gpu/drm/i915/intel_bios.c @@ -1187,6 +1187,15 @@ static void parse_ddi_port(struct drm_i915_private *dev_priv, enum port port, if (is_dvi) { info->alternate_ddc_pin = ddc_pin; + /* + * All VBTs that we got so far for B Stepping has this + * information wrong for Port D. So, let's just ignore for now. + */ + if (IS_CNL_REVID(dev_priv, CNL_REVID_B0, CNL_REVID_B0) && + port == PORT_D) { + info->alternate_ddc_pin = 0; + } + sanitize_ddc_pin(dev_priv, port); } diff --git a/drivers/gpu/drm/i915/intel_cdclk.c b/drivers/gpu/drm/i915/intel_cdclk.c index b8914db7d2e1..1241e5891b29 100644 --- a/drivers/gpu/drm/i915/intel_cdclk.c +++ b/drivers/gpu/drm/i915/intel_cdclk.c @@ -491,6 +491,14 @@ static void vlv_set_cdclk(struct drm_i915_private *dev_priv, int cdclk = cdclk_state->cdclk; u32 val, cmd; + /* There are cases where we can end up here with power domains + * off and a CDCLK frequency other than the minimum, like when + * issuing a modeset without actually changing any display after + * a system suspend. So grab the PIPE-A domain, which covers + * the HW blocks needed for the following programming. + */ + intel_display_power_get(dev_priv, POWER_DOMAIN_PIPE_A); + if (cdclk >= 320000) /* jump to highest voltage for 400MHz too */ cmd = 2; else if (cdclk == 266667) @@ -549,6 +557,8 @@ static void vlv_set_cdclk(struct drm_i915_private *dev_priv, intel_update_cdclk(dev_priv); vlv_program_pfi_credits(dev_priv); + + intel_display_power_put(dev_priv, POWER_DOMAIN_PIPE_A); } static void chv_set_cdclk(struct drm_i915_private *dev_priv, @@ -568,6 +578,14 @@ static void chv_set_cdclk(struct drm_i915_private *dev_priv, return; } + /* There are cases where we can end up here with power domains + * off and a CDCLK frequency other than the minimum, like when + * issuing a modeset without actually changing any display after + * a system suspend. So grab the PIPE-A domain, which covers + * the HW blocks needed for the following programming. + */ + intel_display_power_get(dev_priv, POWER_DOMAIN_PIPE_A); + /* * Specs are full of misinformation, but testing on actual * hardware has shown that we just need to write the desired @@ -590,6 +608,8 @@ static void chv_set_cdclk(struct drm_i915_private *dev_priv, intel_update_cdclk(dev_priv); vlv_program_pfi_credits(dev_priv); + + intel_display_power_put(dev_priv, POWER_DOMAIN_PIPE_A); } static int bdw_calc_cdclk(int max_pixclk) diff --git a/drivers/gpu/drm/i915/intel_color.c b/drivers/gpu/drm/i915/intel_color.c index 306c6b06b330..f85d57555957 100644 --- a/drivers/gpu/drm/i915/intel_color.c +++ b/drivers/gpu/drm/i915/intel_color.c @@ -615,7 +615,7 @@ void intel_color_init(struct drm_crtc *crtc) IS_BROXTON(dev_priv)) { dev_priv->display.load_csc_matrix = i9xx_load_csc_matrix; dev_priv->display.load_luts = broadwell_load_luts; - } else if (IS_GEMINILAKE(dev_priv)) { + } else if (IS_GEMINILAKE(dev_priv) || IS_CANNONLAKE(dev_priv)) { dev_priv->display.load_csc_matrix = i9xx_load_csc_matrix; dev_priv->display.load_luts = glk_load_luts; } else { diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index db8093863f0c..efb13582dc73 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -1103,6 +1103,62 @@ static int skl_calc_wrpll_link(struct drm_i915_private *dev_priv, return dco_freq / (p0 * p1 * p2 * 5); } +static int cnl_calc_wrpll_link(struct drm_i915_private *dev_priv, + uint32_t pll_id) +{ + uint32_t cfgcr0, cfgcr1; + uint32_t p0, p1, p2, dco_freq, ref_clock; + + cfgcr0 = I915_READ(CNL_DPLL_CFGCR0(pll_id)); + cfgcr1 = I915_READ(CNL_DPLL_CFGCR1(pll_id)); + + p0 = cfgcr1 & DPLL_CFGCR1_PDIV_MASK; + p2 = cfgcr1 & DPLL_CFGCR1_KDIV_MASK; + + if (cfgcr1 & DPLL_CFGCR1_QDIV_MODE(1)) + p1 = (cfgcr1 & DPLL_CFGCR1_QDIV_RATIO_MASK) >> + DPLL_CFGCR1_QDIV_RATIO_SHIFT; + else + p1 = 1; + + + switch (p0) { + case DPLL_CFGCR1_PDIV_2: + p0 = 2; + break; + case DPLL_CFGCR1_PDIV_3: + p0 = 3; + break; + case DPLL_CFGCR1_PDIV_5: + p0 = 5; + break; + case DPLL_CFGCR1_PDIV_7: + p0 = 7; + break; + } + + switch (p2) { + case DPLL_CFGCR1_KDIV_1: + p2 = 1; + break; + case DPLL_CFGCR1_KDIV_2: + p2 = 2; + break; + case DPLL_CFGCR1_KDIV_4: + p2 = 4; + break; + } + + ref_clock = dev_priv->cdclk.hw.ref; + + dco_freq = (cfgcr0 & DPLL_CFGCR0_DCO_INTEGER_MASK) * ref_clock; + + dco_freq += (((cfgcr0 & DPLL_CFGCR0_DCO_FRACTION_MASK) >> + DPLL_CFGCR0_DCO_FRAC_SHIFT) * ref_clock) / 0x8000; + + return dco_freq / (p0 * p1 * p2 * 5); +} + static void ddi_dotclock_get(struct intel_crtc_state *pipe_config) { int dotclock; @@ -1124,6 +1180,59 @@ static void ddi_dotclock_get(struct intel_crtc_state *pipe_config) pipe_config->base.adjusted_mode.crtc_clock = dotclock; } +static void cnl_ddi_clock_get(struct intel_encoder *encoder, + struct intel_crtc_state *pipe_config) +{ + struct drm_i915_private *dev_priv = to_i915(encoder->base.dev); + int link_clock = 0; + uint32_t cfgcr0, pll_id; + + pll_id = intel_get_shared_dpll_id(dev_priv, pipe_config->shared_dpll); + + cfgcr0 = I915_READ(CNL_DPLL_CFGCR0(pll_id)); + + if (cfgcr0 & DPLL_CFGCR0_HDMI_MODE) { + link_clock = cnl_calc_wrpll_link(dev_priv, pll_id); + } else { + link_clock = cfgcr0 & DPLL_CFGCR0_LINK_RATE_MASK; + + switch (link_clock) { + case DPLL_CFGCR0_LINK_RATE_810: + link_clock = 81000; + break; + case DPLL_CFGCR0_LINK_RATE_1080: + link_clock = 108000; + break; + case DPLL_CFGCR0_LINK_RATE_1350: + link_clock = 135000; + break; + case DPLL_CFGCR0_LINK_RATE_1620: + link_clock = 162000; + break; + case DPLL_CFGCR0_LINK_RATE_2160: + link_clock = 216000; + break; + case DPLL_CFGCR0_LINK_RATE_2700: + link_clock = 270000; + break; + case DPLL_CFGCR0_LINK_RATE_3240: + link_clock = 324000; + break; + case DPLL_CFGCR0_LINK_RATE_4050: + link_clock = 405000; + break; + default: + WARN(1, "Unsupported link rate\n"); + break; + } + link_clock *= 2; + } + + pipe_config->port_clock = link_clock; + + ddi_dotclock_get(pipe_config); +} + static void skl_ddi_clock_get(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config) { @@ -1267,6 +1376,8 @@ void intel_ddi_clock_get(struct intel_encoder *encoder, skl_ddi_clock_get(encoder, pipe_config); else if (IS_GEN9_LP(dev_priv)) bxt_ddi_clock_get(encoder, pipe_config); + else if (IS_CANNONLAKE(dev_priv)) + cnl_ddi_clock_get(encoder, pipe_config); } void intel_ddi_set_pipe_settings(const struct intel_crtc_state *crtc_state) @@ -1813,11 +1924,14 @@ static void cnl_ddi_vswing_program(struct drm_i915_private *dev_priv, /* Set PORT_TX_DW5 Scaling Mode Sel to 010b. */ val = I915_READ(CNL_PORT_TX_DW5_LN0(port)); + val &= ~SCALING_MODE_SEL_MASK; val |= SCALING_MODE_SEL(2); I915_WRITE(CNL_PORT_TX_DW5_GRP(port), val); /* Program PORT_TX_DW2 */ val = I915_READ(CNL_PORT_TX_DW2_LN0(port)); + val &= ~(SWING_SEL_LOWER_MASK | SWING_SEL_UPPER_MASK | + RCOMP_SCALAR_MASK); val |= SWING_SEL_UPPER(ddi_translations[level].dw2_swing_sel); val |= SWING_SEL_LOWER(ddi_translations[level].dw2_swing_sel); /* Rcomp scalar is fixed as 0x98 for every table entry */ @@ -1828,6 +1942,8 @@ static void cnl_ddi_vswing_program(struct drm_i915_private *dev_priv, /* We cannot write to GRP. It would overrite individual loadgen */ for (ln = 0; ln < 4; ln++) { val = I915_READ(CNL_PORT_TX_DW4_LN(port, ln)); + val &= ~(POST_CURSOR_1_MASK | POST_CURSOR_2_MASK | + CURSOR_COEFF_MASK); val |= POST_CURSOR_1(ddi_translations[level].dw4_post_cursor_1); val |= POST_CURSOR_2(ddi_translations[level].dw4_post_cursor_2); val |= CURSOR_COEFF(ddi_translations[level].dw4_cursor_coeff); @@ -1837,12 +1953,14 @@ static void cnl_ddi_vswing_program(struct drm_i915_private *dev_priv, /* Program PORT_TX_DW5 */ /* All DW5 values are fixed for every table entry */ val = I915_READ(CNL_PORT_TX_DW5_LN0(port)); + val &= ~RTERM_SELECT_MASK; val |= RTERM_SELECT(6); val |= TAP3_DISABLE; I915_WRITE(CNL_PORT_TX_DW5_GRP(port), val); /* Program PORT_TX_DW7 */ val = I915_READ(CNL_PORT_TX_DW7_LN0(port)); + val &= ~N_SCALAR_MASK; val |= N_SCALAR(ddi_translations[level].dw7_n_scalar); I915_WRITE(CNL_PORT_TX_DW7_GRP(port), val); } @@ -1861,9 +1979,12 @@ static void cnl_ddi_vswing_sequence(struct intel_encoder *encoder, u32 level) if ((intel_dp) && (type == INTEL_OUTPUT_EDP || type == INTEL_OUTPUT_DP)) { width = intel_dp->lane_count; rate = intel_dp->link_rate; - } else { + } else if (type == INTEL_OUTPUT_HDMI) { width = 4; /* Rate is always < than 6GHz for HDMI */ + } else { + MISSING_CASE(type); + return; } /* diff --git a/drivers/gpu/drm/i915/intel_device_info.c b/drivers/gpu/drm/i915/intel_device_info.c index 77d3214e1a77..5f91ddc78c7a 100644 --- a/drivers/gpu/drm/i915/intel_device_info.c +++ b/drivers/gpu/drm/i915/intel_device_info.c @@ -363,7 +363,7 @@ void intel_device_info_runtime_init(struct drm_i915_private *dev_priv) */ if (fuse_strap & ILK_INTERNAL_DISPLAY_DISABLE || sfuse_strap & SFUSE_STRAP_DISPLAY_DISABLED || - (dev_priv->pch_type == PCH_CPT && + (HAS_PCH_CPT(dev_priv) && !(sfuse_strap & SFUSE_STRAP_FUSE_LOCK))) { DRM_INFO("Display fused off, disabling\n"); info->num_pipes = 0; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index da5d49407594..e92fd14c06c7 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -3311,7 +3311,7 @@ u32 skl_plane_ctl(const struct intel_crtc_state *crtc_state, plane_ctl = PLANE_CTL_ENABLE; - if (!IS_GEMINILAKE(dev_priv)) { + if (!IS_GEMINILAKE(dev_priv) && !IS_CANNONLAKE(dev_priv)) { plane_ctl |= PLANE_CTL_PIPE_GAMMA_ENABLE | PLANE_CTL_PIPE_CSC_ENABLE | @@ -3367,7 +3367,7 @@ static void skylake_update_primary_plane(struct intel_plane *plane, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - if (IS_GEMINILAKE(dev_priv)) { + if (IS_GEMINILAKE(dev_priv) || IS_CANNONLAKE(dev_priv)) { I915_WRITE_FW(PLANE_COLOR_CTL(pipe, plane_id), PLANE_COLOR_PIPE_GAMMA_ENABLE | PLANE_COLOR_PIPE_CSC_ENABLE | @@ -4612,6 +4612,9 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach, &crtc_state->scaler_state; struct intel_crtc *intel_crtc = to_intel_crtc(crtc_state->base.crtc); + struct drm_i915_private *dev_priv = to_i915(intel_crtc->base.dev); + const struct drm_display_mode *adjusted_mode = + &crtc_state->base.adjusted_mode; int need_scaling; /* @@ -4622,6 +4625,18 @@ skl_update_scaler(struct intel_crtc_state *crtc_state, bool force_detach, need_scaling = src_w != dst_w || src_h != dst_h; /* + * Scaling/fitting not supported in IF-ID mode in GEN9+ + * TODO: Interlace fetch mode doesn't support YUV420 planar formats. + * Once NV12 is enabled, handle it here while allocating scaler + * for NV12. + */ + if (INTEL_GEN(dev_priv) >= 9 && crtc_state->base.enable && + need_scaling && adjusted_mode->flags & DRM_MODE_FLAG_INTERLACE) { + DRM_DEBUG_KMS("Pipe/Plane scaling not supported with IF-ID mode\n"); + return -EINVAL; + } + + /* * if plane is being disabled or scaler is no more required or force detach * - free scaler binded to this plane/crtc * - in order to do this, update crtc->scaler_usage @@ -14773,6 +14788,17 @@ static void quirk_backlight_present(struct drm_device *dev) DRM_INFO("applying backlight present quirk\n"); } +/* Toshiba Satellite P50-C-18C requires T12 delay to be min 800ms + * which is 300 ms greater than eDP spec T12 min. + */ +static void quirk_increase_t12_delay(struct drm_device *dev) +{ + struct drm_i915_private *dev_priv = to_i915(dev); + + dev_priv->quirks |= QUIRK_INCREASE_T12_DELAY; + DRM_INFO("Applying T12 delay quirk\n"); +} + struct intel_quirk { int device; int subsystem_vendor; @@ -14856,6 +14882,9 @@ static struct intel_quirk intel_quirks[] = { /* Dell Chromebook 11 (2015 version) */ { 0x0a16, 0x1028, 0x0a35, quirk_backlight_present }, + + /* Toshiba Satellite P50-C-18C */ + { 0x191B, 0x1179, 0xF840, quirk_increase_t12_delay }, }; static void intel_init_quirks(struct drm_device *dev) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 64fa774c855b..2d42d09428c9 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -4418,8 +4418,6 @@ static bool ibx_digital_port_connected(struct drm_i915_private *dev_priv, u32 bit; switch (port->port) { - case PORT_A: - return true; case PORT_B: bit = SDE_PORTB_HOTPLUG; break; @@ -4443,8 +4441,6 @@ static bool cpt_digital_port_connected(struct drm_i915_private *dev_priv, u32 bit; switch (port->port) { - case PORT_A: - return true; case PORT_B: bit = SDE_PORTB_HOTPLUG_CPT; break; @@ -4454,12 +4450,28 @@ static bool cpt_digital_port_connected(struct drm_i915_private *dev_priv, case PORT_D: bit = SDE_PORTD_HOTPLUG_CPT; break; + default: + MISSING_CASE(port->port); + return false; + } + + return I915_READ(SDEISR) & bit; +} + +static bool spt_digital_port_connected(struct drm_i915_private *dev_priv, + struct intel_digital_port *port) +{ + u32 bit; + + switch (port->port) { + case PORT_A: + bit = SDE_PORTA_HOTPLUG_SPT; + break; case PORT_E: bit = SDE_PORTE_HOTPLUG_SPT; break; default: - MISSING_CASE(port->port); - return false; + return cpt_digital_port_connected(dev_priv, port); } return I915_READ(SDEISR) & bit; @@ -4511,6 +4523,42 @@ static bool gm45_digital_port_connected(struct drm_i915_private *dev_priv, return I915_READ(PORT_HOTPLUG_STAT) & bit; } +static bool ilk_digital_port_connected(struct drm_i915_private *dev_priv, + struct intel_digital_port *port) +{ + if (port->port == PORT_A) + return I915_READ(DEISR) & DE_DP_A_HOTPLUG; + else + return ibx_digital_port_connected(dev_priv, port); +} + +static bool snb_digital_port_connected(struct drm_i915_private *dev_priv, + struct intel_digital_port *port) +{ + if (port->port == PORT_A) + return I915_READ(DEISR) & DE_DP_A_HOTPLUG; + else + return cpt_digital_port_connected(dev_priv, port); +} + +static bool ivb_digital_port_connected(struct drm_i915_private *dev_priv, + struct intel_digital_port *port) +{ + if (port->port == PORT_A) + return I915_READ(DEISR) & DE_DP_A_HOTPLUG_IVB; + else + return cpt_digital_port_connected(dev_priv, port); +} + +static bool bdw_digital_port_connected(struct drm_i915_private *dev_priv, + struct intel_digital_port *port) +{ + if (port->port == PORT_A) + return I915_READ(GEN8_DE_PORT_ISR) & GEN8_PORT_DP_A_HOTPLUG; + else + return cpt_digital_port_connected(dev_priv, port); +} + static bool bxt_digital_port_connected(struct drm_i915_private *dev_priv, struct intel_digital_port *intel_dig_port) { @@ -4547,16 +4595,25 @@ static bool bxt_digital_port_connected(struct drm_i915_private *dev_priv, bool intel_digital_port_connected(struct drm_i915_private *dev_priv, struct intel_digital_port *port) { - if (HAS_PCH_IBX(dev_priv)) - return ibx_digital_port_connected(dev_priv, port); - else if (HAS_PCH_SPLIT(dev_priv)) - return cpt_digital_port_connected(dev_priv, port); + if (HAS_GMCH_DISPLAY(dev_priv)) { + if (IS_GM45(dev_priv)) + return gm45_digital_port_connected(dev_priv, port); + else + return g4x_digital_port_connected(dev_priv, port); + } + + if (IS_GEN5(dev_priv)) + return ilk_digital_port_connected(dev_priv, port); + else if (IS_GEN6(dev_priv)) + return snb_digital_port_connected(dev_priv, port); + else if (IS_GEN7(dev_priv)) + return ivb_digital_port_connected(dev_priv, port); + else if (IS_GEN8(dev_priv)) + return bdw_digital_port_connected(dev_priv, port); else if (IS_GEN9_LP(dev_priv)) return bxt_digital_port_connected(dev_priv, port); - else if (IS_GM45(dev_priv)) - return gm45_digital_port_connected(dev_priv, port); else - return g4x_digital_port_connected(dev_priv, port); + return spt_digital_port_connected(dev_priv, port); } static struct edid * @@ -5121,12 +5178,8 @@ intel_pps_readout_hw_state(struct drm_i915_private *dev_priv, PANEL_POWER_DOWN_DELAY_SHIFT; if (IS_GEN9_LP(dev_priv) || HAS_PCH_CNP(dev_priv)) { - u16 tmp = (pp_ctl & BXT_POWER_CYCLE_DELAY_MASK) >> - BXT_POWER_CYCLE_DELAY_SHIFT; - if (tmp > 0) - seq->t11_t12 = (tmp - 1) * 1000; - else - seq->t11_t12 = 0; + seq->t11_t12 = ((pp_ctl & BXT_POWER_CYCLE_DELAY_MASK) >> + BXT_POWER_CYCLE_DELAY_SHIFT) * 1000; } else { seq->t11_t12 = ((pp_div & PANEL_POWER_CYCLE_DELAY_MASK) >> PANEL_POWER_CYCLE_DELAY_SHIFT) * 1000; @@ -5177,6 +5230,21 @@ intel_dp_init_panel_power_sequencer(struct drm_device *dev, intel_pps_dump_state("cur", &cur); vbt = dev_priv->vbt.edp.pps; + /* On Toshiba Satellite P50-C-18C system the VBT T12 delay + * of 500ms appears to be too short. Ocassionally the panel + * just fails to power back on. Increasing the delay to 800ms + * seems sufficient to avoid this problem. + */ + if (dev_priv->quirks & QUIRK_INCREASE_T12_DELAY) { + vbt.t11_t12 = max_t(u16, vbt.t11_t12, 800 * 10); + DRM_DEBUG_KMS("Increasing T12 panel delay as per the quirk to %d\n", + vbt.t11_t12); + } + /* T11_T12 delay is special and actually in units of 100ms, but zero + * based in the hw (so we need to add 100 ms). But the sw vbt + * table multiplies it with 1000 to make it in units of 100usec, + * too. */ + vbt.t11_t12 += 100 * 10; /* Upper limits from eDP 1.3 spec. Note that we use the clunky units of * our hw here, which are all in 100usec. */ @@ -5280,7 +5348,7 @@ intel_dp_init_panel_power_sequencer_registers(struct drm_device *dev, if (IS_GEN9_LP(dev_priv) || HAS_PCH_CNP(dev_priv)) { pp_div = I915_READ(regs.pp_ctrl); pp_div &= ~BXT_POWER_CYCLE_DELAY_MASK; - pp_div |= (DIV_ROUND_UP((seq->t11_t12 + 1), 1000) + pp_div |= (DIV_ROUND_UP(seq->t11_t12, 1000) << BXT_POWER_CYCLE_DELAY_SHIFT); } else { pp_div = ((100 * div)/2 - 1) << PP_REFERENCE_DIVIDER_SHIFT; diff --git a/drivers/gpu/drm/i915/intel_dp_aux_backlight.c b/drivers/gpu/drm/i915/intel_dp_aux_backlight.c index 228ca06d9f0b..b25cd88fc1c5 100644 --- a/drivers/gpu/drm/i915/intel_dp_aux_backlight.c +++ b/drivers/gpu/drm/i915/intel_dp_aux_backlight.c @@ -98,13 +98,105 @@ intel_dp_aux_set_backlight(const struct drm_connector_state *conn_state, u32 lev } } +/* + * Set PWM Frequency divider to match desired frequency in vbt. + * The PWM Frequency is calculated as 27Mhz / (F x P). + * - Where F = PWM Frequency Pre-Divider value programmed by field 7:0 of the + * EDP_BACKLIGHT_FREQ_SET register (DPCD Address 00728h) + * - Where P = 2^Pn, where Pn is the value programmed by field 4:0 of the + * EDP_PWMGEN_BIT_COUNT register (DPCD Address 00724h) + */ +static bool intel_dp_aux_set_pwm_freq(struct intel_connector *connector) +{ + struct drm_i915_private *dev_priv = to_i915(connector->base.dev); + struct intel_dp *intel_dp = enc_to_intel_dp(&connector->encoder->base); + int freq, fxp, fxp_min, fxp_max, fxp_actual, f = 1; + u8 pn, pn_min, pn_max; + + /* Find desired value of (F x P) + * Note that, if F x P is out of supported range, the maximum value or + * minimum value will applied automatically. So no need to check that. + */ + freq = dev_priv->vbt.backlight.pwm_freq_hz; + DRM_DEBUG_KMS("VBT defined backlight frequency %u Hz\n", freq); + if (!freq) { + DRM_DEBUG_KMS("Use panel default backlight frequency\n"); + return false; + } + + fxp = DIV_ROUND_CLOSEST(KHz(DP_EDP_BACKLIGHT_FREQ_BASE_KHZ), freq); + + /* Use highest possible value of Pn for more granularity of brightness + * adjustment while satifying the conditions below. + * - Pn is in the range of Pn_min and Pn_max + * - F is in the range of 1 and 255 + * - FxP is within 25% of desired value. + * Note: 25% is arbitrary value and may need some tweak. + */ + if (drm_dp_dpcd_readb(&intel_dp->aux, + DP_EDP_PWMGEN_BIT_COUNT_CAP_MIN, &pn_min) != 1) { + DRM_DEBUG_KMS("Failed to read pwmgen bit count cap min\n"); + return false; + } + if (drm_dp_dpcd_readb(&intel_dp->aux, + DP_EDP_PWMGEN_BIT_COUNT_CAP_MAX, &pn_max) != 1) { + DRM_DEBUG_KMS("Failed to read pwmgen bit count cap max\n"); + return false; + } + pn_min &= DP_EDP_PWMGEN_BIT_COUNT_MASK; + pn_max &= DP_EDP_PWMGEN_BIT_COUNT_MASK; + + fxp_min = DIV_ROUND_CLOSEST(fxp * 3, 4); + fxp_max = DIV_ROUND_CLOSEST(fxp * 5, 4); + if (fxp_min < (1 << pn_min) || (255 << pn_max) < fxp_max) { + DRM_DEBUG_KMS("VBT defined backlight frequency out of range\n"); + return false; + } + + for (pn = pn_max; pn >= pn_min; pn--) { + f = clamp(DIV_ROUND_CLOSEST(fxp, 1 << pn), 1, 255); + fxp_actual = f << pn; + if (fxp_min <= fxp_actual && fxp_actual <= fxp_max) + break; + } + + if (drm_dp_dpcd_writeb(&intel_dp->aux, + DP_EDP_PWMGEN_BIT_COUNT, pn) < 0) { + DRM_DEBUG_KMS("Failed to write aux pwmgen bit count\n"); + return false; + } + if (drm_dp_dpcd_writeb(&intel_dp->aux, + DP_EDP_BACKLIGHT_FREQ_SET, (u8) f) < 0) { + DRM_DEBUG_KMS("Failed to write aux backlight freq\n"); + return false; + } + return true; +} + +/* +* Set minimum / maximum dynamic brightness percentage. This value is expressed +* as the percentage of normal brightness in 5% increments. +*/ +static bool +intel_dp_aux_set_dynamic_backlight_percent(struct intel_dp *intel_dp, + u32 min, u32 max) +{ + u8 dbc[] = { DIV_ROUND_CLOSEST(min, 5), DIV_ROUND_CLOSEST(max, 5) }; + + if (drm_dp_dpcd_write(&intel_dp->aux, DP_EDP_DBC_MINIMUM_BRIGHTNESS_SET, + dbc, sizeof(dbc)) < 0) { + DRM_DEBUG_KMS("Failed to write aux DBC brightness level\n"); + return false; + } + return true; +} + static void intel_dp_aux_enable_backlight(const struct intel_crtc_state *crtc_state, const struct drm_connector_state *conn_state) { struct intel_connector *connector = to_intel_connector(conn_state->connector); struct intel_dp *intel_dp = enc_to_intel_dp(&connector->encoder->base); - uint8_t dpcd_buf = 0; - uint8_t edp_backlight_mode = 0; + uint8_t dpcd_buf, new_dpcd_buf, edp_backlight_mode; if (drm_dp_dpcd_readb(&intel_dp->aux, DP_EDP_BACKLIGHT_MODE_SET_REGISTER, &dpcd_buf) != 1) { @@ -113,18 +205,15 @@ static void intel_dp_aux_enable_backlight(const struct intel_crtc_state *crtc_st return; } + new_dpcd_buf = dpcd_buf; edp_backlight_mode = dpcd_buf & DP_EDP_BACKLIGHT_CONTROL_MODE_MASK; switch (edp_backlight_mode) { case DP_EDP_BACKLIGHT_CONTROL_MODE_PWM: case DP_EDP_BACKLIGHT_CONTROL_MODE_PRESET: case DP_EDP_BACKLIGHT_CONTROL_MODE_PRODUCT: - dpcd_buf &= ~DP_EDP_BACKLIGHT_CONTROL_MODE_MASK; - dpcd_buf |= DP_EDP_BACKLIGHT_CONTROL_MODE_DPCD; - if (drm_dp_dpcd_writeb(&intel_dp->aux, - DP_EDP_BACKLIGHT_MODE_SET_REGISTER, dpcd_buf) < 0) { - DRM_DEBUG_KMS("Failed to write aux backlight mode\n"); - } + new_dpcd_buf &= ~DP_EDP_BACKLIGHT_CONTROL_MODE_MASK; + new_dpcd_buf |= DP_EDP_BACKLIGHT_CONTROL_MODE_DPCD; break; /* Do nothing when it is already DPCD mode */ @@ -133,6 +222,25 @@ static void intel_dp_aux_enable_backlight(const struct intel_crtc_state *crtc_st break; } + if (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_FREQ_AUX_SET_CAP) + if (intel_dp_aux_set_pwm_freq(connector)) + new_dpcd_buf |= DP_EDP_BACKLIGHT_FREQ_AUX_SET_ENABLE; + + if (i915.enable_dbc && + (intel_dp->edp_dpcd[2] & DP_EDP_DYNAMIC_BACKLIGHT_CAP)) { + if(intel_dp_aux_set_dynamic_backlight_percent(intel_dp, 0, 100)) { + new_dpcd_buf |= DP_EDP_DYNAMIC_BACKLIGHT_ENABLE; + DRM_DEBUG_KMS("Enable dynamic brightness.\n"); + } + } + + if (new_dpcd_buf != dpcd_buf) { + if (drm_dp_dpcd_writeb(&intel_dp->aux, + DP_EDP_BACKLIGHT_MODE_SET_REGISTER, new_dpcd_buf) < 0) { + DRM_DEBUG_KMS("Failed to write aux backlight mode\n"); + } + } + set_aux_backlight_enable(intel_dp, true); intel_dp_aux_set_backlight(conn_state, connector->panel.backlight.level); } @@ -169,15 +277,66 @@ intel_dp_aux_display_control_capable(struct intel_connector *connector) /* Check the eDP Display control capabilities registers to determine if * the panel can support backlight control over the aux channel */ - if (intel_dp->edp_dpcd[1] & DP_EDP_TCON_BACKLIGHT_ADJUSTMENT_CAP && - (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_AUX_SET_CAP) && - !(intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_PWM_PIN_CAP)) { + if ((intel_dp->edp_dpcd[1] & DP_EDP_TCON_BACKLIGHT_ADJUSTMENT_CAP) && + (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_AUX_SET_CAP)) { DRM_DEBUG_KMS("AUX Backlight Control Supported!\n"); return true; } return false; } +/* + * Heuristic function whether we should use AUX for backlight adjustment or not. + * + * We should use AUX for backlight brightness adjustment if panel doesn't this + * via PWM pin or using AUX is better than using PWM pin. + * + * The heuristic to determine that using AUX pin is better than using PWM pin is + * that the panel support any of the feature list here. + * - Regional backlight brightness adjustment + * - Backlight PWM frequency set + * - More than 8 bits resolution of brightness level + * - Backlight enablement via AUX and not by BL_ENABLE pin + * + * If all above are not true, assume that using PWM pin is better. + */ +static bool +intel_dp_aux_display_control_heuristic(struct intel_connector *connector) +{ + struct intel_dp *intel_dp = enc_to_intel_dp(&connector->encoder->base); + uint8_t reg_val; + + /* Panel doesn't support adjusting backlight brightness via PWN pin */ + if (!(intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_PWM_PIN_CAP)) + return true; + + /* Panel supports regional backlight brightness adjustment */ + if (drm_dp_dpcd_readb(&intel_dp->aux, DP_EDP_GENERAL_CAP_3, + ®_val) != 1) { + DRM_DEBUG_KMS("Failed to read DPCD register 0x%x\n", + DP_EDP_GENERAL_CAP_3); + return false; + } + if (reg_val > 0) + return true; + + /* Panel supports backlight PWM frequency set */ + if (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_FREQ_AUX_SET_CAP) + return true; + + /* Panel supports more than 8 bits resolution of brightness level */ + if (intel_dp->edp_dpcd[2] & DP_EDP_BACKLIGHT_BRIGHTNESS_BYTE_COUNT) + return true; + + /* Panel supports enabling backlight via AUX but not by BL_ENABLE pin */ + if ((intel_dp->edp_dpcd[1] & DP_EDP_BACKLIGHT_AUX_ENABLE_CAP) && + !(intel_dp->edp_dpcd[1] & DP_EDP_BACKLIGHT_PIN_ENABLE_CAP)) + return true; + + return false; + +} + int intel_dp_aux_init_backlight_funcs(struct intel_connector *intel_connector) { struct intel_panel *panel = &intel_connector->panel; @@ -188,6 +347,10 @@ int intel_dp_aux_init_backlight_funcs(struct intel_connector *intel_connector) if (!intel_dp_aux_display_control_capable(intel_connector)) return -ENODEV; + if (i915.enable_dpcd_backlight == -1 && + !intel_dp_aux_display_control_heuristic(intel_connector)) + return -ENODEV; + panel->backlight.setup = intel_dp_aux_setup_backlight; panel->backlight.enable = intel_dp_aux_enable_backlight; panel->backlight.disable = intel_dp_aux_disable_backlight; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index d93efb49a2e2..d17a32437f07 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1858,9 +1858,8 @@ void intel_suspend_gt_powersave(struct drm_i915_private *dev_priv); void gen6_rps_busy(struct drm_i915_private *dev_priv); void gen6_rps_reset_ei(struct drm_i915_private *dev_priv); void gen6_rps_idle(struct drm_i915_private *dev_priv); -void gen6_rps_boost(struct drm_i915_private *dev_priv, - struct intel_rps_client *rps, - unsigned long submitted); +void gen6_rps_boost(struct drm_i915_gem_request *rq, + struct intel_rps_client *rps); void intel_queue_rps_boost_for_request(struct drm_i915_gem_request *req); void g4x_wm_get_hw_state(struct drm_device *dev); void vlv_wm_get_hw_state(struct drm_device *dev); diff --git a/drivers/gpu/drm/i915/intel_engine_cs.c b/drivers/gpu/drm/i915/intel_engine_cs.c index a4487c5b7e37..24db316e0fd1 100644 --- a/drivers/gpu/drm/i915/intel_engine_cs.c +++ b/drivers/gpu/drm/i915/intel_engine_cs.c @@ -149,6 +149,7 @@ __intel_engine_context_size(struct drm_i915_private *dev_priv, u8 class) switch (INTEL_GEN(dev_priv)) { default: MISSING_CASE(INTEL_GEN(dev_priv)); + case 10: case 9: return GEN9_LR_CONTEXT_RENDER_SIZE; case 8: @@ -291,11 +292,9 @@ cleanup: */ int intel_engines_init(struct drm_i915_private *dev_priv) { - struct intel_device_info *device_info = mkwrite_device_info(dev_priv); struct intel_engine_cs *engine; enum intel_engine_id id, err_id; - unsigned int mask = 0; - int err = 0; + int err; for_each_engine(engine, dev_priv, id) { const struct engine_class_info *class_info = @@ -306,40 +305,30 @@ int intel_engines_init(struct drm_i915_private *dev_priv) init = class_info->init_execlists; else init = class_info->init_legacy; - if (!init) { - kfree(engine); - dev_priv->engine[id] = NULL; - continue; - } + + err = -EINVAL; + err_id = id; + + if (GEM_WARN_ON(!init)) + goto cleanup; err = init(engine); - if (err) { - err_id = id; + if (err) goto cleanup; - } GEM_BUG_ON(!engine->submit_request); - mask |= ENGINE_MASK(id); } - /* - * Catch failures to update intel_engines table when the new engines - * are added to the driver by a warning and disabling the forgotten - * engines. - */ - if (WARN_ON(mask != INTEL_INFO(dev_priv)->ring_mask)) - device_info->ring_mask = mask; - - device_info->num_rings = hweight32(mask); - return 0; cleanup: for_each_engine(engine, dev_priv, id) { - if (id >= err_id) + if (id >= err_id) { kfree(engine); - else + dev_priv->engine[id] = NULL; + } else { dev_priv->gt.cleanup_engine(engine); + } } return err; } @@ -821,9 +810,10 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine) I915_WRITE(BDW_SCRATCH1, I915_READ(BDW_SCRATCH1) | GEN9_LBS_SLA_RETRY_TIMER_DECREMENT_ENABLE); - /* WaDisableKillLogic:bxt,skl,kbl,cfl */ - I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | - ECOCHK_DIS_TLB); + /* WaDisableKillLogic:bxt,skl,kbl */ + if (!IS_COFFEELAKE(dev_priv)) + I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | + ECOCHK_DIS_TLB); /* WaClearFlowControlGpgpuContextSave:skl,bxt,kbl,glk,cfl */ /* WaDisablePartialInstShootdown:skl,bxt,kbl,glk,cfl */ @@ -894,10 +884,9 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine) WA_SET_BIT_MASKED(HDC_CHICKEN0, HDC_FORCE_NON_COHERENT); - /* WaDisableHDCInvalidation:skl,bxt,kbl */ - if (!IS_COFFEELAKE(dev_priv)) - I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | - BDW_DISABLE_HDC_INVALIDATION); + /* WaDisableHDCInvalidation:skl,bxt,kbl,cfl */ + I915_WRITE(GAM_ECOCHK, I915_READ(GAM_ECOCHK) | + BDW_DISABLE_HDC_INVALIDATION); /* WaDisableSamplerPowerBypassForSOPingPong:skl,bxt,kbl,cfl */ if (IS_SKYLAKE(dev_priv) || @@ -1340,6 +1329,7 @@ void intel_engines_mark_idle(struct drm_i915_private *i915) for_each_engine(engine, i915, id) { intel_engine_disarm_breadcrumbs(engine); i915_gem_batch_pool_fini(&engine->batch_pool); + tasklet_kill(&engine->irq_tasklet); engine->no_priolist = false; } } diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c index ec6198040381..b953365a3eec 100644 --- a/drivers/gpu/drm/i915/intel_fbdev.c +++ b/drivers/gpu/drm/i915/intel_fbdev.c @@ -542,13 +542,14 @@ static void intel_fbdev_destroy(struct intel_fbdev *ifbdev) drm_fb_helper_fini(&ifbdev->helper); - if (ifbdev->fb) { + if (ifbdev->vma) { mutex_lock(&ifbdev->helper.dev->struct_mutex); intel_unpin_fb_vma(ifbdev->vma); mutex_unlock(&ifbdev->helper.dev->struct_mutex); + } + if (ifbdev->fb) drm_framebuffer_remove(&ifbdev->fb->base); - } kfree(ifbdev); } @@ -772,7 +773,7 @@ void intel_fbdev_set_suspend(struct drm_device *dev, int state, bool synchronous struct intel_fbdev *ifbdev = dev_priv->fbdev; struct fb_info *info; - if (!ifbdev || !ifbdev->fb) + if (!ifbdev || !ifbdev->vma) return; info = ifbdev->helper.fbdev; @@ -819,7 +820,7 @@ void intel_fbdev_output_poll_changed(struct drm_device *dev) { struct intel_fbdev *ifbdev = to_i915(dev)->fbdev; - if (ifbdev && ifbdev->fb) + if (ifbdev) drm_fb_helper_hotplug_event(&ifbdev->helper); } @@ -831,7 +832,7 @@ void intel_fbdev_restore_mode(struct drm_device *dev) return; intel_fbdev_sync(ifbdev); - if (!ifbdev->fb) + if (!ifbdev->vma) return; if (drm_fb_helper_restore_fbdev_mode_unlocked(&ifbdev->helper) == 0) diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c index 7404cf2aac28..699868d81de8 100644 --- a/drivers/gpu/drm/i915/intel_lrc.c +++ b/drivers/gpu/drm/i915/intel_lrc.c @@ -2071,7 +2071,7 @@ void intel_lr_context_resume(struct drm_i915_private *dev_priv) * So to avoid that we reset the context images upon resume. For * simplicity, we just zero everything out. */ - list_for_each_entry(ctx, &dev_priv->context_list, link) { + list_for_each_entry(ctx, &dev_priv->contexts.list, link) { for_each_engine(engine, dev_priv, id) { struct intel_context *ce = &ctx->engine[engine->id]; u32 *reg; diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 48ea0fca1f72..ee2a349cfe68 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3837,7 +3837,7 @@ skl_plane_downscale_amount(const struct intel_crtc_state *cstate, uint_fixed_16_16_t downscale_h, downscale_w; if (WARN_ON(!intel_wm_plane_visible(cstate, pstate))) - return u32_to_fixed_16_16(0); + return u32_to_fixed16(0); /* n.b., src is 16.16 fixed point, dst is whole integer */ if (plane->id == PLANE_CURSOR) { @@ -3861,10 +3861,10 @@ skl_plane_downscale_amount(const struct intel_crtc_state *cstate, dst_h = drm_rect_height(&pstate->base.dst); } - fp_w_ratio = fixed_16_16_div(src_w, dst_w); - fp_h_ratio = fixed_16_16_div(src_h, dst_h); - downscale_w = max_fixed_16_16(fp_w_ratio, u32_to_fixed_16_16(1)); - downscale_h = max_fixed_16_16(fp_h_ratio, u32_to_fixed_16_16(1)); + fp_w_ratio = div_fixed16(src_w, dst_w); + fp_h_ratio = div_fixed16(src_h, dst_h); + downscale_w = max_fixed16(fp_w_ratio, u32_to_fixed16(1)); + downscale_h = max_fixed16(fp_h_ratio, u32_to_fixed16(1)); return mul_fixed16(downscale_w, downscale_h); } @@ -3872,7 +3872,7 @@ skl_plane_downscale_amount(const struct intel_crtc_state *cstate, static uint_fixed_16_16_t skl_pipe_downscale_amount(const struct intel_crtc_state *crtc_state) { - uint_fixed_16_16_t pipe_downscale = u32_to_fixed_16_16(1); + uint_fixed_16_16_t pipe_downscale = u32_to_fixed16(1); if (!crtc_state->base.enable) return pipe_downscale; @@ -3891,10 +3891,10 @@ skl_pipe_downscale_amount(const struct intel_crtc_state *crtc_state) if (!dst_w || !dst_h) return pipe_downscale; - fp_w_ratio = fixed_16_16_div(src_w, dst_w); - fp_h_ratio = fixed_16_16_div(src_h, dst_h); - downscale_w = max_fixed_16_16(fp_w_ratio, u32_to_fixed_16_16(1)); - downscale_h = max_fixed_16_16(fp_h_ratio, u32_to_fixed_16_16(1)); + fp_w_ratio = div_fixed16(src_w, dst_w); + fp_h_ratio = div_fixed16(src_h, dst_h); + downscale_w = max_fixed16(fp_w_ratio, u32_to_fixed16(1)); + downscale_h = max_fixed16(fp_h_ratio, u32_to_fixed16(1)); pipe_downscale = mul_fixed16(downscale_w, downscale_h); } @@ -3913,14 +3913,14 @@ int skl_check_pipe_max_pixel_rate(struct intel_crtc *intel_crtc, int crtc_clock, dotclk; uint32_t pipe_max_pixel_rate; uint_fixed_16_16_t pipe_downscale; - uint_fixed_16_16_t max_downscale = u32_to_fixed_16_16(1); + uint_fixed_16_16_t max_downscale = u32_to_fixed16(1); if (!cstate->base.enable) return 0; drm_atomic_crtc_state_for_each_plane_state(plane, pstate, crtc_state) { uint_fixed_16_16_t plane_downscale; - uint_fixed_16_16_t fp_9_div_8 = fixed_16_16_div(9, 8); + uint_fixed_16_16_t fp_9_div_8 = div_fixed16(9, 8); int bpp; if (!intel_wm_plane_visible(cstate, @@ -3938,7 +3938,7 @@ int skl_check_pipe_max_pixel_rate(struct intel_crtc *intel_crtc, plane_downscale = mul_fixed16(plane_downscale, fp_9_div_8); - max_downscale = max_fixed_16_16(plane_downscale, max_downscale); + max_downscale = max_fixed16(plane_downscale, max_downscale); } pipe_downscale = skl_pipe_downscale_amount(cstate); @@ -4276,7 +4276,7 @@ static uint_fixed_16_16_t skl_wm_method1(uint32_t pixel_rate, uint8_t cpp, return FP_16_16_MAX; wm_intermediate_val = latency * pixel_rate * cpp; - ret = fixed_16_16_div_u64(wm_intermediate_val, 1000 * 512); + ret = div_fixed16(wm_intermediate_val, 1000 * 512); return ret; } @@ -4294,7 +4294,7 @@ static uint_fixed_16_16_t skl_wm_method2(uint32_t pixel_rate, wm_intermediate_val = latency * pixel_rate; wm_intermediate_val = DIV_ROUND_UP(wm_intermediate_val, pipe_htotal * 1000); - ret = mul_u32_fixed_16_16(wm_intermediate_val, plane_blocks_per_line); + ret = mul_u32_fixed16(wm_intermediate_val, plane_blocks_per_line); return ret; } @@ -4306,15 +4306,15 @@ intel_get_linetime_us(struct intel_crtc_state *cstate) uint_fixed_16_16_t linetime_us; if (!cstate->base.active) - return u32_to_fixed_16_16(0); + return u32_to_fixed16(0); pixel_rate = cstate->pixel_rate; if (WARN_ON(pixel_rate == 0)) - return u32_to_fixed_16_16(0); + return u32_to_fixed16(0); crtc_htotal = cstate->base.adjusted_mode.crtc_htotal; - linetime_us = fixed_16_16_div_u64(crtc_htotal * 1000, pixel_rate); + linetime_us = div_fixed16(crtc_htotal * 1000, pixel_rate); return linetime_us; } @@ -4361,7 +4361,7 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, uint32_t plane_bytes_per_line; uint32_t res_blocks, res_lines; uint8_t cpp; - uint32_t width = 0, height = 0; + uint32_t width = 0; uint32_t plane_pixel_rate; uint_fixed_16_16_t y_tile_minimum; uint32_t y_min_scanlines; @@ -4390,7 +4390,6 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, if (plane->id == PLANE_CURSOR) { width = intel_pstate->base.crtc_w; - height = intel_pstate->base.crtc_h; } else { /* * Src coordinates are already rotated by 270 degrees for @@ -4398,16 +4397,13 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, * GTT mapping), hence no need to account for rotation here. */ width = drm_rect_width(&intel_pstate->base.src) >> 16; - height = drm_rect_height(&intel_pstate->base.src) >> 16; } - cpp = fb->format->cpp[0]; + cpp = (fb->format->format == DRM_FORMAT_NV12) ? fb->format->cpp[1] : + fb->format->cpp[0]; plane_pixel_rate = skl_adjusted_plane_pixel_rate(cstate, intel_pstate); if (drm_rotation_90_or_270(pstate->rotation)) { - int cpp = (fb->format->format == DRM_FORMAT_NV12) ? - fb->format->cpp[1] : - fb->format->cpp[0]; switch (cpp) { case 1: @@ -4434,14 +4430,14 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, if (y_tiled) { interm_pbpl = DIV_ROUND_UP(plane_bytes_per_line * y_min_scanlines, 512); - plane_blocks_per_line = fixed_16_16_div(interm_pbpl, + plane_blocks_per_line = div_fixed16(interm_pbpl, y_min_scanlines); } else if (x_tiled) { interm_pbpl = DIV_ROUND_UP(plane_bytes_per_line, 512); - plane_blocks_per_line = u32_to_fixed_16_16(interm_pbpl); + plane_blocks_per_line = u32_to_fixed16(interm_pbpl); } else { interm_pbpl = DIV_ROUND_UP(plane_bytes_per_line, 512) + 1; - plane_blocks_per_line = u32_to_fixed_16_16(interm_pbpl); + plane_blocks_per_line = u32_to_fixed16(interm_pbpl); } method1 = skl_wm_method1(plane_pixel_rate, cpp, latency); @@ -4450,35 +4446,35 @@ static int skl_compute_plane_wm(const struct drm_i915_private *dev_priv, latency, plane_blocks_per_line); - y_tile_minimum = mul_u32_fixed_16_16(y_min_scanlines, - plane_blocks_per_line); + y_tile_minimum = mul_u32_fixed16(y_min_scanlines, + plane_blocks_per_line); if (y_tiled) { - selected_result = max_fixed_16_16(method2, y_tile_minimum); + selected_result = max_fixed16(method2, y_tile_minimum); } else { uint32_t linetime_us; - linetime_us = fixed_16_16_to_u32_round_up( + linetime_us = fixed16_to_u32_round_up( intel_get_linetime_us(cstate)); if ((cpp * cstate->base.adjusted_mode.crtc_htotal / 512 < 1) && (plane_bytes_per_line / 512 < 1)) selected_result = method2; else if ((ddb_allocation && ddb_allocation / - fixed_16_16_to_u32_round_up(plane_blocks_per_line)) >= 1) - selected_result = min_fixed_16_16(method1, method2); + fixed16_to_u32_round_up(plane_blocks_per_line)) >= 1) + selected_result = min_fixed16(method1, method2); else if (latency >= linetime_us) - selected_result = min_fixed_16_16(method1, method2); + selected_result = min_fixed16(method1, method2); else selected_result = method1; } - res_blocks = fixed_16_16_to_u32_round_up(selected_result) + 1; + res_blocks = fixed16_to_u32_round_up(selected_result) + 1; res_lines = div_round_up_fixed16(selected_result, plane_blocks_per_line); if (level >= 1 && level <= 7) { if (y_tiled) { - res_blocks += fixed_16_16_to_u32_round_up(y_tile_minimum); + res_blocks += fixed16_to_u32_round_up(y_tile_minimum); res_lines += y_min_scanlines; } else { res_blocks++; @@ -4563,8 +4559,7 @@ skl_compute_linetime_wm(struct intel_crtc_state *cstate) if (is_fixed16_zero(linetime_us)) return 0; - linetime_wm = fixed_16_16_to_u32_round_up(mul_u32_fixed_16_16(8, - linetime_us)); + linetime_wm = fixed16_to_u32_round_up(mul_u32_fixed16(8, linetime_us)); /* Display WA #1135: bxt. */ if (IS_BROXTON(dev_priv) && dev_priv->ipc_enabled) @@ -5852,7 +5847,7 @@ static u32 intel_rps_limits(struct drm_i915_private *dev_priv, u8 val) * the hw runs at the minimal clock before selecting the desired * frequency, if the down threshold expires in that window we will not * receive a down interrupt. */ - if (IS_GEN9(dev_priv)) { + if (INTEL_GEN(dev_priv) >= 9) { limits = (dev_priv->rps.max_freq_softlimit) << 23; if (val <= dev_priv->rps.min_freq_softlimit) limits |= (dev_priv->rps.min_freq_softlimit) << 14; @@ -5994,7 +5989,7 @@ static int gen6_set_rps(struct drm_i915_private *dev_priv, u8 val) if (val != dev_priv->rps.cur_freq) { gen6_set_rps_thresholds(dev_priv, val); - if (IS_GEN9(dev_priv)) + if (INTEL_GEN(dev_priv) >= 9) I915_WRITE(GEN6_RPNSWREQ, GEN9_FREQUENCY(val)); else if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv)) @@ -6126,47 +6121,35 @@ void gen6_rps_idle(struct drm_i915_private *dev_priv) gen6_sanitize_rps_pm_mask(dev_priv, ~0)); } mutex_unlock(&dev_priv->rps.hw_lock); - - spin_lock(&dev_priv->rps.client_lock); - while (!list_empty(&dev_priv->rps.clients)) - list_del_init(dev_priv->rps.clients.next); - spin_unlock(&dev_priv->rps.client_lock); } -void gen6_rps_boost(struct drm_i915_private *dev_priv, - struct intel_rps_client *rps, - unsigned long submitted) +void gen6_rps_boost(struct drm_i915_gem_request *rq, + struct intel_rps_client *rps) { + struct drm_i915_private *i915 = rq->i915; + bool boost; + /* This is intentionally racy! We peek at the state here, then * validate inside the RPS worker. */ - if (!(dev_priv->gt.awake && - dev_priv->rps.enabled && - dev_priv->rps.cur_freq < dev_priv->rps.boost_freq)) + if (!i915->rps.enabled) return; - /* Force a RPS boost (and don't count it against the client) if - * the GPU is severely congested. - */ - if (rps && time_after(jiffies, submitted + DRM_I915_THROTTLE_JIFFIES)) - rps = NULL; - - spin_lock(&dev_priv->rps.client_lock); - if (rps == NULL || list_empty(&rps->link)) { - spin_lock_irq(&dev_priv->irq_lock); - if (dev_priv->rps.interrupts_enabled) { - dev_priv->rps.client_boost = true; - schedule_work(&dev_priv->rps.work); - } - spin_unlock_irq(&dev_priv->irq_lock); - - if (rps != NULL) { - list_add(&rps->link, &dev_priv->rps.clients); - rps->boosts++; - } else - dev_priv->rps.boosts++; + boost = false; + spin_lock_irq(&rq->lock); + if (!rq->waitboost && !i915_gem_request_completed(rq)) { + atomic_inc(&i915->rps.num_waiters); + rq->waitboost = true; + boost = true; } - spin_unlock(&dev_priv->rps.client_lock); + spin_unlock_irq(&rq->lock); + if (!boost) + return; + + if (READ_ONCE(i915->rps.cur_freq) < i915->rps.boost_freq) + schedule_work(&i915->rps.work); + + atomic_inc(rps ? &rps->boosts : &i915->rps.boosts); } int intel_set_rps(struct drm_i915_private *dev_priv, u8 val) @@ -6365,7 +6348,7 @@ static void gen6_init_rps_frequencies(struct drm_i915_private *dev_priv) dev_priv->rps.efficient_freq = dev_priv->rps.rp1_freq; if (IS_HASWELL(dev_priv) || IS_BROADWELL(dev_priv) || - IS_GEN9_BC(dev_priv)) { + IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) { u32 ddcc_status = 0; if (sandybridge_pcode_read(dev_priv, @@ -6378,7 +6361,7 @@ static void gen6_init_rps_frequencies(struct drm_i915_private *dev_priv) dev_priv->rps.max_freq); } - if (IS_GEN9_BC(dev_priv)) { + if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) { /* Store the frequency values in 16.66 MHZ units, which is * the natural hardware unit for SKL */ @@ -6684,7 +6667,7 @@ static void gen6_update_ring_freq(struct drm_i915_private *dev_priv) /* convert DDR frequency from units of 266.6MHz to bandwidth */ min_ring_freq = mult_frac(min_ring_freq, 8, 3); - if (IS_GEN9_BC(dev_priv)) { + if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) { /* Convert GT frequency to 50 HZ units */ min_gpu_freq = dev_priv->rps.min_freq / GEN9_FREQ_SCALER; max_gpu_freq = dev_priv->rps.max_freq / GEN9_FREQ_SCALER; @@ -6702,7 +6685,7 @@ static void gen6_update_ring_freq(struct drm_i915_private *dev_priv) int diff = max_gpu_freq - gpu_freq; unsigned int ia_freq = 0, ring_freq = 0; - if (IS_GEN9_BC(dev_priv)) { + if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) { /* * ring_freq = 2 * GT. ring_freq is in 100MHz units * No floor required for ring frequency on SKL. @@ -7833,7 +7816,7 @@ void intel_enable_gt_powersave(struct drm_i915_private *dev_priv) } else if (INTEL_GEN(dev_priv) >= 9) { gen9_enable_rc6(dev_priv); gen9_enable_rps(dev_priv); - if (IS_GEN9_BC(dev_priv)) + if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) gen6_update_ring_freq(dev_priv); } else if (IS_BROADWELL(dev_priv)) { gen8_enable_rps(dev_priv); @@ -9078,7 +9061,7 @@ static int chv_freq_opcode(struct drm_i915_private *dev_priv, int val) int intel_gpu_freq(struct drm_i915_private *dev_priv, int val) { - if (IS_GEN9(dev_priv)) + if (INTEL_GEN(dev_priv) >= 9) return DIV_ROUND_CLOSEST(val * GT_FREQUENCY_MULTIPLIER, GEN9_FREQ_SCALER); else if (IS_CHERRYVIEW(dev_priv)) @@ -9091,7 +9074,7 @@ int intel_gpu_freq(struct drm_i915_private *dev_priv, int val) int intel_freq_opcode(struct drm_i915_private *dev_priv, int val) { - if (IS_GEN9(dev_priv)) + if (INTEL_GEN(dev_priv) >= 9) return DIV_ROUND_CLOSEST(val * GEN9_FREQ_SCALER, GT_FREQUENCY_MULTIPLIER); else if (IS_CHERRYVIEW(dev_priv)) @@ -9113,7 +9096,7 @@ static void __intel_rps_boost_work(struct work_struct *work) struct drm_i915_gem_request *req = boost->req; if (!i915_gem_request_completed(req)) - gen6_rps_boost(req->i915, NULL, req->emitted_jiffies); + gen6_rps_boost(req, NULL); i915_gem_request_put(req); kfree(boost); @@ -9142,11 +9125,10 @@ void intel_queue_rps_boost_for_request(struct drm_i915_gem_request *req) void intel_pm_setup(struct drm_i915_private *dev_priv) { mutex_init(&dev_priv->rps.hw_lock); - spin_lock_init(&dev_priv->rps.client_lock); INIT_DELAYED_WORK(&dev_priv->rps.autoenable_work, __intel_autoenable_gt_powersave); - INIT_LIST_HEAD(&dev_priv->rps.clients); + atomic_set(&dev_priv->rps.num_waiters, 0); dev_priv->pm.suspended = false; atomic_set(&dev_priv->pm.wakeref_count, 0); diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index acd1da9b62a3..5224b7abb8a3 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -2140,7 +2140,7 @@ static void intel_ring_default_vfuncs(struct drm_i915_private *dev_priv, engine->emit_breadcrumb = gen6_sema_emit_breadcrumb; - num_rings = hweight32(INTEL_INFO(dev_priv)->ring_mask) - 1; + num_rings = INTEL_INFO(dev_priv)->num_rings - 1; if (INTEL_GEN(dev_priv) >= 8) { engine->emit_breadcrumb_sz += num_rings * 6; } else { @@ -2184,8 +2184,7 @@ int intel_init_render_ring_buffer(struct intel_engine_cs *engine) engine->semaphore.signal = gen8_rcs_signal; - num_rings = - hweight32(INTEL_INFO(dev_priv)->ring_mask) - 1; + num_rings = INTEL_INFO(dev_priv)->num_rings - 1; engine->emit_breadcrumb_sz += num_rings * 8; } } else if (INTEL_GEN(dev_priv) >= 6) { diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h index 6aa20ac8cde3..d33c93444c0d 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.h +++ b/drivers/gpu/drm/i915/intel_ringbuffer.h @@ -121,6 +121,7 @@ struct intel_engine_hangcheck { unsigned long action_timestamp; int deadlock; struct intel_instdone instdone; + struct drm_i915_gem_request *active_request; bool stalled; }; diff --git a/drivers/gpu/drm/i915/intel_runtime_pm.c b/drivers/gpu/drm/i915/intel_runtime_pm.c index efe80ed5fd4d..f630d632a976 100644 --- a/drivers/gpu/drm/i915/intel_runtime_pm.c +++ b/drivers/gpu/drm/i915/intel_runtime_pm.c @@ -341,6 +341,59 @@ static void skl_power_well_pre_disable(struct drm_i915_private *dev_priv, 1 << PIPE_C | 1 << PIPE_B); } +static void gen9_wait_for_power_well_enable(struct drm_i915_private *dev_priv, + struct i915_power_well *power_well) +{ + int id = power_well->id; + + /* Timeout for PW1:10 us, AUX:not specified, other PWs:20 us. */ + WARN_ON(intel_wait_for_register(dev_priv, + HSW_PWR_WELL_DRIVER, + SKL_POWER_WELL_STATE(id), + SKL_POWER_WELL_STATE(id), + 1)); +} + +static u32 gen9_power_well_requesters(struct drm_i915_private *dev_priv, int id) +{ + u32 req_mask = SKL_POWER_WELL_REQ(id); + u32 ret; + + ret = I915_READ(HSW_PWR_WELL_BIOS) & req_mask ? 1 : 0; + ret |= I915_READ(HSW_PWR_WELL_DRIVER) & req_mask ? 2 : 0; + ret |= I915_READ(HSW_PWR_WELL_KVMR) & req_mask ? 4 : 0; + ret |= I915_READ(HSW_PWR_WELL_DEBUG) & req_mask ? 8 : 0; + + return ret; +} + +static void gen9_wait_for_power_well_disable(struct drm_i915_private *dev_priv, + struct i915_power_well *power_well) +{ + int id = power_well->id; + bool disabled; + u32 reqs; + + /* + * Bspec doesn't require waiting for PWs to get disabled, but still do + * this for paranoia. The known cases where a PW will be forced on: + * - a KVMR request on any power well via the KVMR request register + * - a DMC request on PW1 and MISC_IO power wells via the BIOS and + * DEBUG request registers + * Skip the wait in case any of the request bits are set and print a + * diagnostic message. + */ + wait_for((disabled = !(I915_READ(HSW_PWR_WELL_DRIVER) & + SKL_POWER_WELL_STATE(id))) || + (reqs = gen9_power_well_requesters(dev_priv, id)), 1); + if (disabled) + return; + + DRM_DEBUG_KMS("%s forced on (bios:%d driver:%d kvmr:%d debug:%d)\n", + power_well->name, + !!(reqs & 1), !!(reqs & 2), !!(reqs & 4), !!(reqs & 8)); +} + static void hsw_set_power_well(struct drm_i915_private *dev_priv, struct i915_power_well *power_well, bool enable) { @@ -549,7 +602,9 @@ static void assert_can_enable_dc9(struct drm_i915_private *dev_priv) "DC9 already programmed to be enabled.\n"); WARN_ONCE(I915_READ(DC_STATE_EN) & DC_STATE_EN_UPTO_DC5, "DC5 still not disabled to enable DC9.\n"); - WARN_ONCE(I915_READ(HSW_PWR_WELL_DRIVER), "Power well on.\n"); + WARN_ONCE(I915_READ(HSW_PWR_WELL_DRIVER) & + SKL_POWER_WELL_REQ(SKL_DISP_PW_2), + "Power well 2 on.\n"); WARN_ONCE(intel_irqs_enabled(dev_priv), "Interrupts not disabled yet.\n"); @@ -744,45 +799,6 @@ void skl_disable_dc6(struct drm_i915_private *dev_priv) gen9_set_dc_state(dev_priv, DC_STATE_DISABLE); } -static void -gen9_sanitize_power_well_requests(struct drm_i915_private *dev_priv, - struct i915_power_well *power_well) -{ - enum skl_disp_power_wells power_well_id = power_well->id; - u32 val; - u32 mask; - - mask = SKL_POWER_WELL_REQ(power_well_id); - - val = I915_READ(HSW_PWR_WELL_KVMR); - if (WARN_ONCE(val & mask, "Clearing unexpected KVMR request for %s\n", - power_well->name)) - I915_WRITE(HSW_PWR_WELL_KVMR, val & ~mask); - - val = I915_READ(HSW_PWR_WELL_BIOS); - val |= I915_READ(HSW_PWR_WELL_DEBUG); - - if (!(val & mask)) - return; - - /* - * DMC is known to force on the request bits for power well 1 on SKL - * and BXT and the misc IO power well on SKL but we don't expect any - * other request bits to be set, so WARN for those. - */ - if (power_well_id == SKL_DISP_PW_1 || - (IS_GEN9_BC(dev_priv) && - power_well_id == SKL_DISP_PW_MISC_IO)) - DRM_DEBUG_DRIVER("Clearing auxiliary requests for %s forced on " - "by DMC\n", power_well->name); - else - WARN_ONCE(1, "Clearing unexpected auxiliary requests for %s\n", - power_well->name); - - I915_WRITE(HSW_PWR_WELL_BIOS, val & ~mask); - I915_WRITE(HSW_PWR_WELL_DEBUG, val & ~mask); -} - static void skl_set_power_well(struct drm_i915_private *dev_priv, struct i915_power_well *power_well, bool enable) { @@ -846,6 +862,8 @@ static void skl_set_power_well(struct drm_i915_private *dev_priv, DRM_DEBUG_KMS("Enabling %s\n", power_well->name); check_fuse_status = true; } + + gen9_wait_for_power_well_enable(dev_priv, power_well); } else { if (enable_requested) { I915_WRITE(HSW_PWR_WELL_DRIVER, tmp & ~req_mask); @@ -853,14 +871,9 @@ static void skl_set_power_well(struct drm_i915_private *dev_priv, DRM_DEBUG_KMS("Disabling %s\n", power_well->name); } - gen9_sanitize_power_well_requests(dev_priv, power_well); + gen9_wait_for_power_well_disable(dev_priv, power_well); } - if (wait_for(!!(I915_READ(HSW_PWR_WELL_DRIVER) & state_mask) == enable, - 1)) - DRM_ERROR("%s %s timeout\n", - power_well->name, enable ? "enable" : "disable"); - if (check_fuse_status) { if (power_well->id == SKL_DISP_PW_1) { if (intel_wait_for_register(dev_priv, @@ -2479,7 +2492,7 @@ static uint32_t get_allowed_dc_mask(const struct drm_i915_private *dev_priv, int requested_dc; int max_dc; - if (IS_GEN9_BC(dev_priv)) { + if (IS_GEN9_BC(dev_priv) || IS_CANNONLAKE(dev_priv)) { max_dc = 2; mask = 0; } else if (IS_GEN9_LP(dev_priv)) { @@ -2694,13 +2707,18 @@ static void skl_display_core_uninit(struct drm_i915_private *dev_priv) mutex_lock(&power_domains->lock); - well = lookup_power_well(dev_priv, SKL_DISP_PW_MISC_IO); - intel_power_well_disable(dev_priv, well); - + /* + * BSpec says to keep the MISC IO power well enabled here, only + * remove our request for power well 1. + * Note that even though the driver's request is removed power well 1 + * may stay enabled after this due to DMC's own request on it. + */ well = lookup_power_well(dev_priv, SKL_DISP_PW_1); intel_power_well_disable(dev_priv, well); mutex_unlock(&power_domains->lock); + + usleep_range(10, 30); /* 10 us delay per Bspec */ } void bxt_display_core_init(struct drm_i915_private *dev_priv, @@ -2751,13 +2769,19 @@ void bxt_display_core_uninit(struct drm_i915_private *dev_priv) /* The spec doesn't call for removing the reset handshake flag */ - /* Disable PG1 */ + /* + * Disable PW1 (PG1). + * Note that even though the driver's request is removed power well 1 + * may stay enabled after this due to DMC's own request on it. + */ mutex_lock(&power_domains->lock); well = lookup_power_well(dev_priv, SKL_DISP_PW_1); intel_power_well_disable(dev_priv, well); mutex_unlock(&power_domains->lock); + + usleep_range(10, 30); /* 10 us delay per Bspec */ } #define CNL_PROCMON_IDX(val) \ @@ -2821,7 +2845,10 @@ static void cnl_display_core_init(struct drm_i915_private *dev_priv, bool resume val |= CL_POWER_DOWN_ENABLE; I915_WRITE(CNL_PORT_CL1CM_DW5, val); - /* 4. Enable Power Well 1 (PG1) and Aux IO Power */ + /* + * 4. Enable Power Well 1 (PG1). + * The AUX IO power wells will be enabled on demand. + */ mutex_lock(&power_domains->lock); well = lookup_power_well(dev_priv, SKL_DISP_PW_1); intel_power_well_enable(dev_priv, well); @@ -2853,12 +2880,18 @@ static void cnl_display_core_uninit(struct drm_i915_private *dev_priv) /* 3. Disable CD clock */ cnl_uninit_cdclk(dev_priv); - /* 4. Disable Power Well 1 (PG1) and Aux IO Power */ + /* + * 4. Disable Power Well 1 (PG1). + * The AUX IO power wells are toggled on demand, so they are already + * disabled at this point. + */ mutex_lock(&power_domains->lock); well = lookup_power_well(dev_priv, SKL_DISP_PW_1); intel_power_well_disable(dev_priv, well); mutex_unlock(&power_domains->lock); + usleep_range(10, 30); /* 10 us delay per Bspec */ + /* 5. Disable Comp */ val = I915_READ(CHICKEN_MISC_2); val |= COMP_PWR_DOWN; diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index 85d9ff361e74..e58a47db9a9d 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -1344,7 +1344,7 @@ static void intel_sdvo_pre_enable(struct intel_encoder *intel_encoder, sdvox |= (9 << 19) | SDVO_BORDER_ENABLE; } - if (INTEL_PCH_TYPE(dev_priv) >= PCH_CPT) + if (HAS_PCH_CPT(dev_priv)) sdvox |= SDVO_PIPE_SEL_CPT(crtc->pipe); else sdvox |= SDVO_PIPE_SEL(crtc->pipe); diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 0c650c2cbca8..94f9a1332dbf 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -262,7 +262,7 @@ skl_update_plane(struct intel_plane *plane, spin_lock_irqsave(&dev_priv->uncore.lock, irqflags); - if (IS_GEMINILAKE(dev_priv)) { + if (IS_GEMINILAKE(dev_priv) || IS_CANNONLAKE(dev_priv)) { I915_WRITE_FW(PLANE_COLOR_CTL(pipe, plane_id), PLANE_COLOR_PIPE_GAMMA_ENABLE | PLANE_COLOR_PIPE_CSC_ENABLE | diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c index 9882724bc2b6..deb4430541cf 100644 --- a/drivers/gpu/drm/i915/intel_uncore.c +++ b/drivers/gpu/drm/i915/intel_uncore.c @@ -643,7 +643,7 @@ find_fw_domain(struct drm_i915_private *dev_priv, u32 offset) { .start = (s), .end = (e), .domains = (d) } #define HAS_FWTABLE(dev_priv) \ - (IS_GEN9(dev_priv) || \ + (INTEL_GEN(dev_priv) >= 9 || \ IS_CHERRYVIEW(dev_priv) || \ IS_VALLEYVIEW(dev_priv)) @@ -1072,7 +1072,7 @@ static void intel_uncore_fw_domains_init(struct drm_i915_private *dev_priv) dev_priv->uncore.fw_clear = _MASKED_BIT_DISABLE(FORCEWAKE_KERNEL); } - if (IS_GEN9(dev_priv)) { + if (INTEL_GEN(dev_priv) >= 9) { dev_priv->uncore.funcs.force_wake_get = fw_domains_get; dev_priv->uncore.funcs.force_wake_put = fw_domains_put; fw_domain_init(dev_priv, FW_DOMAIN_ID_RENDER, @@ -1719,6 +1719,17 @@ bool intel_has_gpu_reset(struct drm_i915_private *dev_priv) return intel_get_gpu_reset(dev_priv) != NULL; } +/* + * When GuC submission is enabled, GuC manages ELSP and can initiate the + * engine reset too. For now, fall back to full GPU reset if it is enabled. + */ +bool intel_has_reset_engine(struct drm_i915_private *dev_priv) +{ + return (dev_priv->info.has_reset_engine && + !dev_priv->guc.execbuf_client && + i915.reset >= 2); +} + int intel_guc_reset(struct drm_i915_private *dev_priv) { int ret; diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c b/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c index d15cc9d3a5cd..89dc25a5a53b 100644 --- a/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c +++ b/drivers/gpu/drm/i915/selftests/i915_gem_dmabuf.c @@ -246,9 +246,9 @@ static int igt_dmabuf_export_vmap(void *arg) i915_gem_object_put(obj); ptr = dma_buf_vmap(dmabuf); - if (IS_ERR(ptr)) { - err = PTR_ERR(ptr); - pr_err("dma_buf_vmap failed with err=%d\n", err); + if (!ptr) { + pr_err("dma_buf_vmap failed\n"); + err = -ENOMEM; goto out; } diff --git a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c index 50710e3f1caa..6b132caffa18 100644 --- a/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/selftests/i915_gem_gtt.c @@ -197,6 +197,9 @@ static int lowlevel_hole(struct drm_i915_private *i915, { I915_RND_STATE(seed_prng); unsigned int size; + struct i915_vma mock_vma; + + memset(&mock_vma, 0, sizeof(struct i915_vma)); /* Keep creating larger objects until one cannot fit into the hole */ for (size = 12; (hole_end - hole_start) >> size; size++) { @@ -255,8 +258,11 @@ static int lowlevel_hole(struct drm_i915_private *i915, vm->allocate_va_range(vm, addr, BIT_ULL(size))) break; - vm->insert_entries(vm, obj->mm.pages, addr, - I915_CACHE_NONE, 0); + mock_vma.pages = obj->mm.pages; + mock_vma.node.size = BIT_ULL(size); + mock_vma.node.start = addr; + + vm->insert_entries(vm, &mock_vma, I915_CACHE_NONE, 0); } count = n; diff --git a/drivers/gpu/drm/i915/selftests/i915_vma.c b/drivers/gpu/drm/i915/selftests/i915_vma.c index fb9072d5877f..2e86ec136b35 100644 --- a/drivers/gpu/drm/i915/selftests/i915_vma.c +++ b/drivers/gpu/drm/i915/selftests/i915_vma.c @@ -186,16 +186,20 @@ static int igt_vma_create(void *arg) goto end; } - list_for_each_entry_safe(ctx, cn, &contexts, link) + list_for_each_entry_safe(ctx, cn, &contexts, link) { + list_del_init(&ctx->link); mock_context_close(ctx); + } } end: /* Final pass to lookup all created contexts */ err = create_vmas(i915, &objects, &contexts); out: - list_for_each_entry_safe(ctx, cn, &contexts, link) + list_for_each_entry_safe(ctx, cn, &contexts, link) { + list_del_init(&ctx->link); mock_context_close(ctx); + } list_for_each_entry_safe(obj, on, &objects, st_link) i915_gem_object_put(obj); diff --git a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c index aa31d6c0cdfb..7096c3911cd3 100644 --- a/drivers/gpu/drm/i915/selftests/intel_hangcheck.c +++ b/drivers/gpu/drm/i915/selftests/intel_hangcheck.c @@ -316,6 +316,56 @@ static int igt_global_reset(void *arg) GEM_BUG_ON(test_bit(I915_RESET_HANDOFF, &i915->gpu_error.flags)); clear_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); + wake_up_all(&i915->gpu_error.reset_queue); + + if (i915_terminally_wedged(&i915->gpu_error)) + err = -EIO; + + return err; +} + +static int igt_reset_engine(void *arg) +{ + struct drm_i915_private *i915 = arg; + struct intel_engine_cs *engine; + enum intel_engine_id id; + unsigned int reset_count, reset_engine_count; + int err = 0; + + /* Check that we can issue a global GPU and engine reset */ + + if (!intel_has_reset_engine(i915)) + return 0; + + for_each_engine(engine, i915, id) { + set_bit(I915_RESET_ENGINE + engine->id, &i915->gpu_error.flags); + reset_count = i915_reset_count(&i915->gpu_error); + reset_engine_count = i915_reset_engine_count(&i915->gpu_error, + engine); + + err = i915_reset_engine(engine); + if (err) { + pr_err("i915_reset_engine failed\n"); + break; + } + + if (i915_reset_count(&i915->gpu_error) != reset_count) { + pr_err("Full GPU reset recorded! (engine reset expected)\n"); + err = -EINVAL; + break; + } + + if (i915_reset_engine_count(&i915->gpu_error, engine) == + reset_engine_count) { + pr_err("No %s engine reset recorded!\n", engine->name); + err = -EINVAL; + break; + } + + clear_bit(I915_RESET_ENGINE + engine->id, + &i915->gpu_error.flags); + } + if (i915_terminally_wedged(&i915->gpu_error)) err = -EIO; @@ -404,6 +454,7 @@ fini: unlock: mutex_unlock(&i915->drm.struct_mutex); clear_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); + wake_up_all(&i915->gpu_error.reset_queue); if (i915_terminally_wedged(&i915->gpu_error)) return -EIO; @@ -519,11 +570,117 @@ fini: unlock: mutex_unlock(&i915->drm.struct_mutex); clear_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); + wake_up_all(&i915->gpu_error.reset_queue); + + if (i915_terminally_wedged(&i915->gpu_error)) + return -EIO; + + return err; +} + +static int igt_render_engine_reset_fallback(void *arg) +{ + struct drm_i915_private *i915 = arg; + struct intel_engine_cs *engine = i915->engine[RCS]; + struct hang h; + struct drm_i915_gem_request *rq; + unsigned int reset_count, reset_engine_count; + int err = 0; + + /* Check that we can issue a global GPU and engine reset */ + + if (!intel_has_reset_engine(i915)) + return 0; + + set_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); + mutex_lock(&i915->drm.struct_mutex); + + err = hang_init(&h, i915); + if (err) + goto err_unlock; + + rq = hang_create_request(&h, engine, i915->kernel_context); + if (IS_ERR(rq)) { + err = PTR_ERR(rq); + goto err_fini; + } + + i915_gem_request_get(rq); + __i915_add_request(rq, true); + + /* make reset engine fail */ + rq->fence.error = -EIO; + + if (!wait_for_hang(&h, rq)) { + pr_err("Failed to start request %x\n", rq->fence.seqno); + err = -EIO; + goto err_request; + } + + reset_engine_count = i915_reset_engine_count(&i915->gpu_error, engine); + reset_count = fake_hangcheck(rq); + + /* unlock since we'll call handle_error */ + mutex_unlock(&i915->drm.struct_mutex); + clear_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); + wake_up_all(&i915->gpu_error.reset_queue); + + i915_handle_error(i915, intel_engine_flag(engine), "live test"); + + if (i915_reset_engine_count(&i915->gpu_error, engine) != + reset_engine_count) { + pr_err("render engine reset recorded! (full reset expected)\n"); + err = -EINVAL; + goto out_rq; + } + + if (i915_reset_count(&i915->gpu_error) == reset_count) { + pr_err("No full GPU reset recorded!\n"); + err = -EINVAL; + goto out_rq; + } + + /* + * by using fence.error = -EIO, full reset sets the wedged flag, do one + * more full reset to re-enable the hw. + */ + if (i915_terminally_wedged(&i915->gpu_error)) { + set_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); + rq->fence.error = 0; + + mutex_lock(&i915->drm.struct_mutex); + set_bit(I915_RESET_HANDOFF, &i915->gpu_error.flags); + i915_reset(i915); + GEM_BUG_ON(test_bit(I915_RESET_HANDOFF, + &i915->gpu_error.flags)); + mutex_unlock(&i915->drm.struct_mutex); + + if (i915_reset_count(&i915->gpu_error) == reset_count) { + pr_err("No full GPU reset recorded!\n"); + err = -EINVAL; + goto out_rq; + } + } + +out_rq: + i915_gem_request_put(rq); + hang_fini(&h); +out_backoff: + clear_bit(I915_RESET_BACKOFF, &i915->gpu_error.flags); + wake_up_all(&i915->gpu_error.reset_queue); if (i915_terminally_wedged(&i915->gpu_error)) return -EIO; return err; + +err_request: + i915_gem_request_put(rq); +err_fini: + hang_fini(&h); +err_unlock: + mutex_unlock(&i915->drm.struct_mutex); + goto out_backoff; } int intel_hangcheck_live_selftests(struct drm_i915_private *i915) @@ -531,8 +688,10 @@ int intel_hangcheck_live_selftests(struct drm_i915_private *i915) static const struct i915_subtest tests[] = { SUBTEST(igt_hang_sanitycheck), SUBTEST(igt_global_reset), + SUBTEST(igt_reset_engine), SUBTEST(igt_wait_reset), SUBTEST(igt_reset_queue), + SUBTEST(igt_render_engine_reset_fallback), }; if (!intel_has_gpu_reset(i915)) diff --git a/drivers/gpu/drm/i915/selftests/mock_context.c b/drivers/gpu/drm/i915/selftests/mock_context.c index f8b9cc212b02..9c7c68181f82 100644 --- a/drivers/gpu/drm/i915/selftests/mock_context.c +++ b/drivers/gpu/drm/i915/selftests/mock_context.c @@ -48,7 +48,7 @@ mock_context(struct drm_i915_private *i915, if (!ctx->vma_lut.ht) goto err_free; - ret = ida_simple_get(&i915->context_hw_ida, + ret = ida_simple_get(&i915->contexts.hw_ida, 0, MAX_CONTEXT_HW_ID, GFP_KERNEL); if (ret < 0) goto err_vma_ht; @@ -86,3 +86,12 @@ void mock_context_close(struct i915_gem_context *ctx) i915_gem_context_put(ctx); } + +void mock_init_contexts(struct drm_i915_private *i915) +{ + INIT_LIST_HEAD(&i915->contexts.list); + ida_init(&i915->contexts.hw_ida); + + INIT_WORK(&i915->contexts.free_work, contexts_free_worker); + init_llist_head(&i915->contexts.free_list); +} diff --git a/drivers/gpu/drm/i915/selftests/mock_context.h b/drivers/gpu/drm/i915/selftests/mock_context.h index 2427e5c0916a..383941a61124 100644 --- a/drivers/gpu/drm/i915/selftests/mock_context.h +++ b/drivers/gpu/drm/i915/selftests/mock_context.h @@ -25,6 +25,8 @@ #ifndef __MOCK_CONTEXT_H #define __MOCK_CONTEXT_H +void mock_init_contexts(struct drm_i915_private *i915); + struct i915_gem_context * mock_context(struct drm_i915_private *i915, const char *name); diff --git a/drivers/gpu/drm/i915/selftests/mock_gem_device.c b/drivers/gpu/drm/i915/selftests/mock_gem_device.c index 627e2aa09766..47613d20bba8 100644 --- a/drivers/gpu/drm/i915/selftests/mock_gem_device.c +++ b/drivers/gpu/drm/i915/selftests/mock_gem_device.c @@ -57,11 +57,12 @@ static void mock_device_release(struct drm_device *dev) cancel_delayed_work_sync(&i915->gt.retire_work); cancel_delayed_work_sync(&i915->gt.idle_work); + flush_workqueue(i915->wq); mutex_lock(&i915->drm.struct_mutex); for_each_engine(engine, i915, id) mock_engine_free(engine); - i915_gem_context_fini(i915); + i915_gem_contexts_fini(i915); mutex_unlock(&i915->drm.struct_mutex); drain_workqueue(i915->wq); @@ -160,7 +161,7 @@ struct drm_i915_private *mock_gem_device(void) INIT_LIST_HEAD(&i915->mm.unbound_list); INIT_LIST_HEAD(&i915->mm.bound_list); - ida_init(&i915->context_hw_ida); + mock_init_contexts(i915); INIT_DELAYED_WORK(&i915->gt.retire_work, mock_retire_work_handler); INIT_DELAYED_WORK(&i915->gt.idle_work, mock_idle_work_handler); diff --git a/drivers/gpu/drm/i915/selftests/mock_gtt.c b/drivers/gpu/drm/i915/selftests/mock_gtt.c index a61309c7cb3e..f2118cf535a0 100644 --- a/drivers/gpu/drm/i915/selftests/mock_gtt.c +++ b/drivers/gpu/drm/i915/selftests/mock_gtt.c @@ -33,8 +33,7 @@ static void mock_insert_page(struct i915_address_space *vm, } static void mock_insert_entries(struct i915_address_space *vm, - struct sg_table *st, - u64 start, + struct i915_vma *vma, enum i915_cache_level level, u32 flags) { } diff --git a/drivers/gpu/drm/mediatek/Makefile b/drivers/gpu/drm/mediatek/Makefile index bf2e5be1ab30..e37b55a23a65 100644 --- a/drivers/gpu/drm/mediatek/Makefile +++ b/drivers/gpu/drm/mediatek/Makefile @@ -1,4 +1,5 @@ -mediatek-drm-y := mtk_disp_ovl.o \ +mediatek-drm-y := mtk_disp_color.o \ + mtk_disp_ovl.o \ mtk_disp_rdma.o \ mtk_drm_crtc.o \ mtk_drm_ddp.o \ diff --git a/drivers/gpu/drm/mediatek/mtk_disp_color.c b/drivers/gpu/drm/mediatek/mtk_disp_color.c new file mode 100644 index 000000000000..ef79a6d55646 --- /dev/null +++ b/drivers/gpu/drm/mediatek/mtk_disp_color.c @@ -0,0 +1,176 @@ +/* + * Copyright (c) 2017 MediaTek Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include <drm/drmP.h> +#include <linux/clk.h> +#include <linux/component.h> +#include <linux/of_device.h> +#include <linux/of_irq.h> +#include <linux/platform_device.h> + +#include "mtk_drm_crtc.h" +#include "mtk_drm_ddp_comp.h" + +#define DISP_COLOR_CFG_MAIN 0x0400 +#define DISP_COLOR_START_MT2701 0x0f00 +#define DISP_COLOR_START_MT8173 0x0c00 +#define DISP_COLOR_START(comp) ((comp)->data->color_offset) +#define DISP_COLOR_WIDTH(comp) (DISP_COLOR_START(comp) + 0x50) +#define DISP_COLOR_HEIGHT(comp) (DISP_COLOR_START(comp) + 0x54) + +#define COLOR_BYPASS_ALL BIT(7) +#define COLOR_SEQ_SEL BIT(13) + +struct mtk_disp_color_data { + unsigned int color_offset; +}; + +/** + * struct mtk_disp_color - DISP_COLOR driver structure + * @ddp_comp - structure containing type enum and hardware resources + * @crtc - associated crtc to report irq events to + */ +struct mtk_disp_color { + struct mtk_ddp_comp ddp_comp; + struct drm_crtc *crtc; + const struct mtk_disp_color_data *data; +}; + +static inline struct mtk_disp_color *comp_to_color(struct mtk_ddp_comp *comp) +{ + return container_of(comp, struct mtk_disp_color, ddp_comp); +} + +static void mtk_color_config(struct mtk_ddp_comp *comp, unsigned int w, + unsigned int h, unsigned int vrefresh, + unsigned int bpc) +{ + struct mtk_disp_color *color = comp_to_color(comp); + + writel(w, comp->regs + DISP_COLOR_WIDTH(color)); + writel(h, comp->regs + DISP_COLOR_HEIGHT(color)); +} + +static void mtk_color_start(struct mtk_ddp_comp *comp) +{ + struct mtk_disp_color *color = comp_to_color(comp); + + writel(COLOR_BYPASS_ALL | COLOR_SEQ_SEL, + comp->regs + DISP_COLOR_CFG_MAIN); + writel(0x1, comp->regs + DISP_COLOR_START(color)); +} + +static const struct mtk_ddp_comp_funcs mtk_disp_color_funcs = { + .config = mtk_color_config, + .start = mtk_color_start, +}; + +static int mtk_disp_color_bind(struct device *dev, struct device *master, + void *data) +{ + struct mtk_disp_color *priv = dev_get_drvdata(dev); + struct drm_device *drm_dev = data; + int ret; + + ret = mtk_ddp_comp_register(drm_dev, &priv->ddp_comp); + if (ret < 0) { + dev_err(dev, "Failed to register component %s: %d\n", + dev->of_node->full_name, ret); + return ret; + } + + return 0; +} + +static void mtk_disp_color_unbind(struct device *dev, struct device *master, + void *data) +{ + struct mtk_disp_color *priv = dev_get_drvdata(dev); + struct drm_device *drm_dev = data; + + mtk_ddp_comp_unregister(drm_dev, &priv->ddp_comp); +} + +static const struct component_ops mtk_disp_color_component_ops = { + .bind = mtk_disp_color_bind, + .unbind = mtk_disp_color_unbind, +}; + +static int mtk_disp_color_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct mtk_disp_color *priv; + int comp_id; + int ret; + + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + comp_id = mtk_ddp_comp_get_id(dev->of_node, MTK_DISP_COLOR); + if (comp_id < 0) { + dev_err(dev, "Failed to identify by alias: %d\n", comp_id); + return comp_id; + } + + ret = mtk_ddp_comp_init(dev, dev->of_node, &priv->ddp_comp, comp_id, + &mtk_disp_color_funcs); + if (ret) { + dev_err(dev, "Failed to initialize component: %d\n", ret); + return ret; + } + + priv->data = of_device_get_match_data(dev); + + platform_set_drvdata(pdev, priv); + + ret = component_add(dev, &mtk_disp_color_component_ops); + if (ret) + dev_err(dev, "Failed to add component: %d\n", ret); + + return ret; +} + +static int mtk_disp_color_remove(struct platform_device *pdev) +{ + component_del(&pdev->dev, &mtk_disp_color_component_ops); + + return 0; +} + +static const struct mtk_disp_color_data mt2701_color_driver_data = { + .color_offset = DISP_COLOR_START_MT2701, +}; + +static const struct mtk_disp_color_data mt8173_color_driver_data = { + .color_offset = DISP_COLOR_START_MT8173, +}; + +static const struct of_device_id mtk_disp_color_driver_dt_match[] = { + { .compatible = "mediatek,mt2701-disp-color", + .data = &mt2701_color_driver_data}, + { .compatible = "mediatek,mt8173-disp-color", + .data = &mt8173_color_driver_data}, + {}, +}; +MODULE_DEVICE_TABLE(of, mtk_disp_color_driver_dt_match); + +struct platform_driver mtk_disp_color_driver = { + .probe = mtk_disp_color_probe, + .remove = mtk_disp_color_remove, + .driver = { + .name = "mediatek-disp-color", + .owner = THIS_MODULE, + .of_match_table = mtk_disp_color_driver_dt_match, + }, +}; diff --git a/drivers/gpu/drm/mediatek/mtk_disp_ovl.c b/drivers/gpu/drm/mediatek/mtk_disp_ovl.c index a14d7d64d7b1..35bc5babdbf7 100644 --- a/drivers/gpu/drm/mediatek/mtk_disp_ovl.c +++ b/drivers/gpu/drm/mediatek/mtk_disp_ovl.c @@ -42,9 +42,12 @@ #define OVL_RDMA_MEM_GMC 0x40402020 #define OVL_CON_BYTE_SWAP BIT(24) +#define OVL_CON_MTX_YUV_TO_RGB (6 << 16) #define OVL_CON_CLRFMT_RGB (1 << 12) #define OVL_CON_CLRFMT_RGBA8888 (2 << 12) #define OVL_CON_CLRFMT_ARGB8888 (3 << 12) +#define OVL_CON_CLRFMT_UYVY (4 << 12) +#define OVL_CON_CLRFMT_YUYV (5 << 12) #define OVL_CON_CLRFMT_RGB565(ovl) ((ovl)->data->fmt_rgb565_is_0 ? \ 0 : OVL_CON_CLRFMT_RGB) #define OVL_CON_CLRFMT_RGB888(ovl) ((ovl)->data->fmt_rgb565_is_0 ? \ @@ -176,6 +179,10 @@ static unsigned int ovl_fmt_convert(struct mtk_disp_ovl *ovl, unsigned int fmt) case DRM_FORMAT_XBGR8888: case DRM_FORMAT_ABGR8888: return OVL_CON_CLRFMT_RGBA8888 | OVL_CON_BYTE_SWAP; + case DRM_FORMAT_UYVY: + return OVL_CON_CLRFMT_UYVY | OVL_CON_MTX_YUV_TO_RGB; + case DRM_FORMAT_YUYV: + return OVL_CON_CLRFMT_YUYV | OVL_CON_MTX_YUV_TO_RGB; } } diff --git a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c index fc65c57dda8c..1f0ef17aa455 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_crtc.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_crtc.c @@ -561,6 +561,8 @@ int mtk_drm_crtc_create(struct drm_device *drm_dev, mtk_crtc->ddp_comp = devm_kmalloc_array(dev, mtk_crtc->ddp_comp_nr, sizeof(*mtk_crtc->ddp_comp), GFP_KERNEL); + if (!mtk_crtc->ddp_comp) + return -ENOMEM; mtk_crtc->mutex = mtk_disp_mutex_get(priv->mutex_dev, pipe); if (IS_ERR(mtk_crtc->mutex)) { diff --git a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c index 8b52416b6e41..07d7ea2268ef 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_ddp_comp.c @@ -38,13 +38,6 @@ #define DISP_REG_UFO_START 0x0000 -#define DISP_COLOR_CFG_MAIN 0x0400 -#define DISP_COLOR_START_MT2701 0x0f00 -#define DISP_COLOR_START_MT8173 0x0c00 -#define DISP_COLOR_START(comp) ((comp)->data->color_offset) -#define DISP_COLOR_WIDTH(comp) (DISP_COLOR_START(comp) + 0x50) -#define DISP_COLOR_HEIGHT(comp) (DISP_COLOR_START(comp) + 0x54) - #define DISP_AAL_EN 0x0000 #define DISP_AAL_SIZE 0x0030 @@ -55,9 +48,6 @@ #define LUT_10BIT_MASK 0x03ff -#define COLOR_BYPASS_ALL BIT(7) -#define COLOR_SEQ_SEL BIT(13) - #define OD_RELAYMODE BIT(0) #define UFO_BYPASS BIT(2) @@ -82,20 +72,6 @@ #define DITHER_ADD_LSHIFT_G(x) (((x) & 0x7) << 4) #define DITHER_ADD_RSHIFT_G(x) (((x) & 0x7) << 0) -struct mtk_disp_color_data { - unsigned int color_offset; -}; - -struct mtk_disp_color { - struct mtk_ddp_comp ddp_comp; - const struct mtk_disp_color_data *data; -}; - -static inline struct mtk_disp_color *comp_to_color(struct mtk_ddp_comp *comp) -{ - return container_of(comp, struct mtk_disp_color, ddp_comp); -} - void mtk_dither_set(struct mtk_ddp_comp *comp, unsigned int bpc, unsigned int CFG) { @@ -119,25 +95,6 @@ void mtk_dither_set(struct mtk_ddp_comp *comp, unsigned int bpc, } } -static void mtk_color_config(struct mtk_ddp_comp *comp, unsigned int w, - unsigned int h, unsigned int vrefresh, - unsigned int bpc) -{ - struct mtk_disp_color *color = comp_to_color(comp); - - writel(w, comp->regs + DISP_COLOR_WIDTH(color)); - writel(h, comp->regs + DISP_COLOR_HEIGHT(color)); -} - -static void mtk_color_start(struct mtk_ddp_comp *comp) -{ - struct mtk_disp_color *color = comp_to_color(comp); - - writel(COLOR_BYPASS_ALL | COLOR_SEQ_SEL, - comp->regs + DISP_COLOR_CFG_MAIN); - writel(0x1, comp->regs + DISP_COLOR_START(color)); -} - static void mtk_od_config(struct mtk_ddp_comp *comp, unsigned int w, unsigned int h, unsigned int vrefresh, unsigned int bpc) @@ -229,11 +186,6 @@ static const struct mtk_ddp_comp_funcs ddp_gamma = { .stop = mtk_gamma_stop, }; -static const struct mtk_ddp_comp_funcs ddp_color = { - .config = mtk_color_config, - .start = mtk_color_start, -}; - static const struct mtk_ddp_comp_funcs ddp_od = { .config = mtk_od_config, .start = mtk_od_start, @@ -268,8 +220,8 @@ struct mtk_ddp_comp_match { static const struct mtk_ddp_comp_match mtk_ddp_matches[DDP_COMPONENT_ID_MAX] = { [DDP_COMPONENT_AAL] = { MTK_DISP_AAL, 0, &ddp_aal }, [DDP_COMPONENT_BLS] = { MTK_DISP_BLS, 0, NULL }, - [DDP_COMPONENT_COLOR0] = { MTK_DISP_COLOR, 0, &ddp_color }, - [DDP_COMPONENT_COLOR1] = { MTK_DISP_COLOR, 1, &ddp_color }, + [DDP_COMPONENT_COLOR0] = { MTK_DISP_COLOR, 0, NULL }, + [DDP_COMPONENT_COLOR1] = { MTK_DISP_COLOR, 1, NULL }, [DDP_COMPONENT_DPI0] = { MTK_DPI, 0, NULL }, [DDP_COMPONENT_DSI0] = { MTK_DSI, 0, NULL }, [DDP_COMPONENT_DSI1] = { MTK_DSI, 1, NULL }, @@ -286,22 +238,6 @@ static const struct mtk_ddp_comp_match mtk_ddp_matches[DDP_COMPONENT_ID_MAX] = { [DDP_COMPONENT_WDMA1] = { MTK_DISP_WDMA, 1, NULL }, }; -static const struct mtk_disp_color_data mt2701_color_driver_data = { - .color_offset = DISP_COLOR_START_MT2701, -}; - -static const struct mtk_disp_color_data mt8173_color_driver_data = { - .color_offset = DISP_COLOR_START_MT8173, -}; - -static const struct of_device_id mtk_disp_color_driver_dt_match[] = { - { .compatible = "mediatek,mt2701-disp-color", - .data = &mt2701_color_driver_data}, - { .compatible = "mediatek,mt8173-disp-color", - .data = &mt8173_color_driver_data}, - {}, -}; - int mtk_ddp_comp_get_id(struct device_node *node, enum mtk_ddp_comp_type comp_type) { @@ -324,23 +260,11 @@ int mtk_ddp_comp_init(struct device *dev, struct device_node *node, enum mtk_ddp_comp_type type; struct device_node *larb_node; struct platform_device *larb_pdev; - const struct of_device_id *match; - struct mtk_disp_color *color; if (comp_id < 0 || comp_id >= DDP_COMPONENT_ID_MAX) return -EINVAL; type = mtk_ddp_matches[comp_id].type; - if (type == MTK_DISP_COLOR) { - devm_kfree(dev, comp); - color = devm_kzalloc(dev, sizeof(*color), GFP_KERNEL); - if (!color) - return -ENOMEM; - - match = of_match_node(mtk_disp_color_driver_dt_match, node); - color->data = match->data; - comp = &color->ddp_comp; - } comp->id = comp_id; comp->funcs = funcs ?: mtk_ddp_matches[comp_id].funcs; diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.c b/drivers/gpu/drm/mediatek/mtk_drm_drv.c index b2596f35104b..c8163525d444 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_drv.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.c @@ -443,11 +443,12 @@ static int mtk_drm_probe(struct platform_device *pdev) private->comp_node[comp_id] = of_node_get(node); /* - * Currently only the OVL, RDMA, DSI, and DPI blocks have + * Currently only the COLOR, OVL, RDMA, DSI, and DPI blocks have * separate component platform drivers and initialize their own * DDP component structure. The others are initialized here. */ - if (comp_type == MTK_DISP_OVL || + if (comp_type == MTK_DISP_COLOR || + comp_type == MTK_DISP_OVL || comp_type == MTK_DISP_RDMA || comp_type == MTK_DSI || comp_type == MTK_DPI) { @@ -570,6 +571,7 @@ static struct platform_driver mtk_drm_platform_driver = { static struct platform_driver * const mtk_drm_drivers[] = { &mtk_ddp_driver, + &mtk_disp_color_driver, &mtk_disp_ovl_driver, &mtk_disp_rdma_driver, &mtk_dpi_driver, @@ -580,33 +582,14 @@ static struct platform_driver * const mtk_drm_drivers[] = { static int __init mtk_drm_init(void) { - int ret; - int i; - - for (i = 0; i < ARRAY_SIZE(mtk_drm_drivers); i++) { - ret = platform_driver_register(mtk_drm_drivers[i]); - if (ret < 0) { - pr_err("Failed to register %s driver: %d\n", - mtk_drm_drivers[i]->driver.name, ret); - goto err; - } - } - - return 0; - -err: - while (--i >= 0) - platform_driver_unregister(mtk_drm_drivers[i]); - - return ret; + return platform_register_drivers(mtk_drm_drivers, + ARRAY_SIZE(mtk_drm_drivers)); } static void __exit mtk_drm_exit(void) { - int i; - - for (i = ARRAY_SIZE(mtk_drm_drivers) - 1; i >= 0; i--) - platform_driver_unregister(mtk_drm_drivers[i]); + platform_unregister_drivers(mtk_drm_drivers, + ARRAY_SIZE(mtk_drm_drivers)); } module_init(mtk_drm_init); diff --git a/drivers/gpu/drm/mediatek/mtk_drm_drv.h b/drivers/gpu/drm/mediatek/mtk_drm_drv.h index aef8747d810b..c3378c452c0a 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_drv.h +++ b/drivers/gpu/drm/mediatek/mtk_drm_drv.h @@ -59,6 +59,7 @@ struct mtk_drm_private { }; extern struct platform_driver mtk_ddp_driver; +extern struct platform_driver mtk_disp_color_driver; extern struct platform_driver mtk_disp_ovl_driver; extern struct platform_driver mtk_disp_rdma_driver; extern struct platform_driver mtk_dpi_driver; diff --git a/drivers/gpu/drm/mediatek/mtk_drm_plane.c b/drivers/gpu/drm/mediatek/mtk_drm_plane.c index e405e89ed5e5..1a59b9ab4aa8 100644 --- a/drivers/gpu/drm/mediatek/mtk_drm_plane.c +++ b/drivers/gpu/drm/mediatek/mtk_drm_plane.c @@ -28,6 +28,8 @@ static const u32 formats[] = { DRM_FORMAT_XRGB8888, DRM_FORMAT_ARGB8888, DRM_FORMAT_RGB565, + DRM_FORMAT_UYVY, + DRM_FORMAT_YUYV, }; static void mtk_plane_reset(struct drm_plane *plane) diff --git a/drivers/gpu/drm/mediatek/mtk_dsi.c b/drivers/gpu/drm/mediatek/mtk_dsi.c index b5cc6e12334c..97253c8f813b 100644 --- a/drivers/gpu/drm/mediatek/mtk_dsi.c +++ b/drivers/gpu/drm/mediatek/mtk_dsi.c @@ -930,7 +930,7 @@ static u32 mtk_dsi_recv_cnt(u8 type, u8 *read_data) DRM_INFO("type is 0x02, try again\n"); break; default: - DRM_INFO("type(0x%x) cannot be non-recognite\n", type); + DRM_INFO("type(0x%x) not recognized\n", type); break; } diff --git a/drivers/gpu/drm/mediatek/mtk_hdmi.c b/drivers/gpu/drm/mediatek/mtk_hdmi.c index 5c0d02444bd3..252d373990bf 100644 --- a/drivers/gpu/drm/mediatek/mtk_hdmi.c +++ b/drivers/gpu/drm/mediatek/mtk_hdmi.c @@ -1778,33 +1778,14 @@ static struct platform_driver * const mtk_hdmi_drivers[] = { static int __init mtk_hdmitx_init(void) { - int ret; - int i; - - for (i = 0; i < ARRAY_SIZE(mtk_hdmi_drivers); i++) { - ret = platform_driver_register(mtk_hdmi_drivers[i]); - if (ret < 0) { - pr_err("Failed to register %s driver: %d\n", - mtk_hdmi_drivers[i]->driver.name, ret); - goto err; - } - } - - return 0; - -err: - while (--i >= 0) - platform_driver_unregister(mtk_hdmi_drivers[i]); - - return ret; + return platform_register_drivers(mtk_hdmi_drivers, + ARRAY_SIZE(mtk_hdmi_drivers)); } static void __exit mtk_hdmitx_exit(void) { - int i; - - for (i = ARRAY_SIZE(mtk_hdmi_drivers) - 1; i >= 0; i--) - platform_driver_unregister(mtk_hdmi_drivers[i]); + platform_unregister_drivers(mtk_hdmi_drivers, + ARRAY_SIZE(mtk_hdmi_drivers)); } module_init(mtk_hdmitx_init); diff --git a/drivers/gpu/drm/mga/mga_drv.h b/drivers/gpu/drm/mga/mga_drv.h index 45cf363d25ad..a45bb22275a7 100644 --- a/drivers/gpu/drm/mga/mga_drv.h +++ b/drivers/gpu/drm/mga/mga_drv.h @@ -159,6 +159,8 @@ extern int mga_dma_bootstrap(struct drm_device *dev, void *data, struct drm_file *file_priv); extern int mga_dma_init(struct drm_device *dev, void *data, struct drm_file *file_priv); +extern int mga_getparam(struct drm_device *dev, void *data, + struct drm_file *file_priv); extern int mga_dma_flush(struct drm_device *dev, void *data, struct drm_file *file_priv); extern int mga_dma_reset(struct drm_device *dev, void *data, diff --git a/drivers/gpu/drm/mga/mga_ioc32.c b/drivers/gpu/drm/mga/mga_ioc32.c index 729bfd56b55f..245fb2e359cf 100644 --- a/drivers/gpu/drm/mga/mga_ioc32.c +++ b/drivers/gpu/drm/mga/mga_ioc32.c @@ -61,46 +61,25 @@ static int compat_mga_init(struct file *file, unsigned int cmd, unsigned long arg) { drm_mga_init32_t init32; - drm_mga_init_t __user *init; - int err = 0, i; + drm_mga_init_t init; if (copy_from_user(&init32, (void __user *)arg, sizeof(init32))) return -EFAULT; - init = compat_alloc_user_space(sizeof(*init)); - if (!access_ok(VERIFY_WRITE, init, sizeof(*init)) - || __put_user(init32.func, &init->func) - || __put_user(init32.sarea_priv_offset, &init->sarea_priv_offset) - || __put_user(init32.chipset, &init->chipset) - || __put_user(init32.sgram, &init->sgram) - || __put_user(init32.maccess, &init->maccess) - || __put_user(init32.fb_cpp, &init->fb_cpp) - || __put_user(init32.front_offset, &init->front_offset) - || __put_user(init32.front_pitch, &init->front_pitch) - || __put_user(init32.back_offset, &init->back_offset) - || __put_user(init32.back_pitch, &init->back_pitch) - || __put_user(init32.depth_cpp, &init->depth_cpp) - || __put_user(init32.depth_offset, &init->depth_offset) - || __put_user(init32.depth_pitch, &init->depth_pitch) - || __put_user(init32.fb_offset, &init->fb_offset) - || __put_user(init32.mmio_offset, &init->mmio_offset) - || __put_user(init32.status_offset, &init->status_offset) - || __put_user(init32.warp_offset, &init->warp_offset) - || __put_user(init32.primary_offset, &init->primary_offset) - || __put_user(init32.buffers_offset, &init->buffers_offset)) - return -EFAULT; - - for (i = 0; i < MGA_NR_TEX_HEAPS; i++) { - err |= - __put_user(init32.texture_offset[i], - &init->texture_offset[i]); - err |= - __put_user(init32.texture_size[i], &init->texture_size[i]); - } - if (err) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_MGA_INIT, (unsigned long)init); + init.func = init32.func; + init.sarea_priv_offset = init32.sarea_priv_offset; + memcpy(&init.chipset, &init32.chipset, + offsetof(drm_mga_init_t, fb_offset) - + offsetof(drm_mga_init_t, chipset)); + init.fb_offset = init32.fb_offset; + init.mmio_offset = init32.mmio_offset; + init.status_offset = init32.status_offset; + init.warp_offset = init32.warp_offset; + init.primary_offset = init32.primary_offset; + init.buffers_offset = init32.buffers_offset; + + return drm_ioctl_kernel(file, mga_dma_init, &init, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); } typedef struct drm_mga_getparam32 { @@ -112,19 +91,14 @@ static int compat_mga_getparam(struct file *file, unsigned int cmd, unsigned long arg) { drm_mga_getparam32_t getparam32; - drm_mga_getparam_t __user *getparam; + drm_mga_getparam_t getparam; if (copy_from_user(&getparam32, (void __user *)arg, sizeof(getparam32))) return -EFAULT; - getparam = compat_alloc_user_space(sizeof(*getparam)); - if (!access_ok(VERIFY_WRITE, getparam, sizeof(*getparam)) - || __put_user(getparam32.param, &getparam->param) - || __put_user((void __user *)(unsigned long)getparam32.value, - &getparam->value)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_MGA_GETPARAM, (unsigned long)getparam); + getparam.param = getparam32.param; + getparam.value = compat_ptr(getparam32.value); + return drm_ioctl_kernel(file, mga_getparam, &getparam, DRM_AUTH); } typedef struct drm_mga_drm_bootstrap32 { @@ -141,48 +115,33 @@ static int compat_mga_dma_bootstrap(struct file *file, unsigned int cmd, unsigned long arg) { drm_mga_dma_bootstrap32_t dma_bootstrap32; - drm_mga_dma_bootstrap_t __user *dma_bootstrap; + drm_mga_dma_bootstrap_t dma_bootstrap; int err; if (copy_from_user(&dma_bootstrap32, (void __user *)arg, sizeof(dma_bootstrap32))) return -EFAULT; - dma_bootstrap = compat_alloc_user_space(sizeof(*dma_bootstrap)); - if (!access_ok(VERIFY_WRITE, dma_bootstrap, sizeof(*dma_bootstrap)) - || __put_user(dma_bootstrap32.texture_handle, - &dma_bootstrap->texture_handle) - || __put_user(dma_bootstrap32.texture_size, - &dma_bootstrap->texture_size) - || __put_user(dma_bootstrap32.primary_size, - &dma_bootstrap->primary_size) - || __put_user(dma_bootstrap32.secondary_bin_count, - &dma_bootstrap->secondary_bin_count) - || __put_user(dma_bootstrap32.secondary_bin_size, - &dma_bootstrap->secondary_bin_size) - || __put_user(dma_bootstrap32.agp_mode, &dma_bootstrap->agp_mode) - || __put_user(dma_bootstrap32.agp_size, &dma_bootstrap->agp_size)) - return -EFAULT; + dma_bootstrap.texture_handle = dma_bootstrap32.texture_handle; + dma_bootstrap.texture_size = dma_bootstrap32.texture_size; + dma_bootstrap.primary_size = dma_bootstrap32.primary_size; + dma_bootstrap.secondary_bin_count = dma_bootstrap32.secondary_bin_count; + dma_bootstrap.secondary_bin_size = dma_bootstrap32.secondary_bin_size; + dma_bootstrap.agp_mode = dma_bootstrap32.agp_mode; + dma_bootstrap.agp_size = dma_bootstrap32.agp_size; - err = drm_ioctl(file, DRM_IOCTL_MGA_DMA_BOOTSTRAP, - (unsigned long)dma_bootstrap); + err = drm_ioctl_kernel(file, mga_dma_bootstrap, &dma_bootstrap, + DRM_AUTH|DRM_MASTER|DRM_ROOT_ONLY); if (err) return err; - if (__get_user(dma_bootstrap32.texture_handle, - &dma_bootstrap->texture_handle) - || __get_user(dma_bootstrap32.texture_size, - &dma_bootstrap->texture_size) - || __get_user(dma_bootstrap32.primary_size, - &dma_bootstrap->primary_size) - || __get_user(dma_bootstrap32.secondary_bin_count, - &dma_bootstrap->secondary_bin_count) - || __get_user(dma_bootstrap32.secondary_bin_size, - &dma_bootstrap->secondary_bin_size) - || __get_user(dma_bootstrap32.agp_mode, &dma_bootstrap->agp_mode) - || __get_user(dma_bootstrap32.agp_size, &dma_bootstrap->agp_size)) - return -EFAULT; - + dma_bootstrap32.texture_handle = dma_bootstrap.texture_handle; + dma_bootstrap32.texture_size = dma_bootstrap.texture_size; + dma_bootstrap32.primary_size = dma_bootstrap.primary_size; + dma_bootstrap32.secondary_bin_count = dma_bootstrap.secondary_bin_count; + dma_bootstrap32.secondary_bin_size = dma_bootstrap.secondary_bin_size; + dma_bootstrap32.agp_mode = dma_bootstrap.agp_mode; + dma_bootstrap32.agp_size = dma_bootstrap.agp_size; if (copy_to_user((void __user *)arg, &dma_bootstrap32, sizeof(dma_bootstrap32))) return -EFAULT; @@ -190,10 +149,14 @@ static int compat_mga_dma_bootstrap(struct file *file, unsigned int cmd, return 0; } -drm_ioctl_compat_t *mga_compat_ioctls[] = { - [DRM_MGA_INIT] = compat_mga_init, - [DRM_MGA_GETPARAM] = compat_mga_getparam, - [DRM_MGA_DMA_BOOTSTRAP] = compat_mga_dma_bootstrap, +static struct { + drm_ioctl_compat_t *fn; + char *name; +} mga_compat_ioctls[] = { +#define DRM_IOCTL32_DEF(n, f)[DRM_##n] = {.fn = f, .name = #n} + DRM_IOCTL32_DEF(MGA_INIT, compat_mga_init), + DRM_IOCTL32_DEF(MGA_GETPARAM, compat_mga_getparam), + DRM_IOCTL32_DEF(MGA_DMA_BOOTSTRAP, compat_mga_dma_bootstrap), }; /** @@ -208,19 +171,27 @@ drm_ioctl_compat_t *mga_compat_ioctls[] = { long mga_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { unsigned int nr = DRM_IOCTL_NR(cmd); + struct drm_file *file_priv = filp->private_data; drm_ioctl_compat_t *fn = NULL; int ret; if (nr < DRM_COMMAND_BASE) return drm_compat_ioctl(filp, cmd, arg); - if (nr < DRM_COMMAND_BASE + ARRAY_SIZE(mga_compat_ioctls)) - fn = mga_compat_ioctls[nr - DRM_COMMAND_BASE]; - - if (fn != NULL) - ret = (*fn) (filp, cmd, arg); - else - ret = drm_ioctl(filp, cmd, arg); - + if (nr >= DRM_COMMAND_BASE + ARRAY_SIZE(mga_compat_ioctls)) + return drm_ioctl(filp, cmd, arg); + + fn = mga_compat_ioctls[nr - DRM_COMMAND_BASE].fn; + if (!fn) + return drm_ioctl(filp, cmd, arg); + + DRM_DEBUG("pid=%d, dev=0x%lx, auth=%d, %s\n", + task_pid_nr(current), + (long)old_encode_dev(file_priv->minor->kdev->devt), + file_priv->authenticated, + mga_compat_ioctls[nr - DRM_COMMAND_BASE].name); + ret = (*fn) (filp, cmd, arg); + if (ret) + DRM_DEBUG("ret = %d\n", ret); return ret; } diff --git a/drivers/gpu/drm/mga/mga_state.c b/drivers/gpu/drm/mga/mga_state.c index 792f924496fc..e5f6b735f575 100644 --- a/drivers/gpu/drm/mga/mga_state.c +++ b/drivers/gpu/drm/mga/mga_state.c @@ -1005,7 +1005,7 @@ static int mga_dma_blit(struct drm_device *dev, void *data, struct drm_file *fil return 0; } -static int mga_getparam(struct drm_device *dev, void *data, struct drm_file *file_priv) +int mga_getparam(struct drm_device *dev, void *data, struct drm_file *file_priv) { drm_mga_private_t *dev_priv = dev->dev_private; drm_mga_getparam_t *param = data; diff --git a/drivers/gpu/drm/nouveau/nouveau_acpi.c b/drivers/gpu/drm/nouveau/nouveau_acpi.c index 39468c218027..7459ef9943ec 100644 --- a/drivers/gpu/drm/nouveau/nouveau_acpi.c +++ b/drivers/gpu/drm/nouveau/nouveau_acpi.c @@ -60,15 +60,13 @@ bool nouveau_is_v1_dsm(void) { } #ifdef CONFIG_VGA_SWITCHEROO -static const char nouveau_dsm_muid[] = { - 0xA0, 0xA0, 0x95, 0x9D, 0x60, 0x00, 0x48, 0x4D, - 0xB3, 0x4D, 0x7E, 0x5F, 0xEA, 0x12, 0x9F, 0xD4, -}; +static const guid_t nouveau_dsm_muid = + GUID_INIT(0x9D95A0A0, 0x0060, 0x4D48, + 0xB3, 0x4D, 0x7E, 0x5F, 0xEA, 0x12, 0x9F, 0xD4); -static const char nouveau_op_dsm_muid[] = { - 0xF8, 0xD8, 0x86, 0xA4, 0xDA, 0x0B, 0x1B, 0x47, - 0xA7, 0x2B, 0x60, 0x42, 0xA6, 0xB5, 0xBE, 0xE0, -}; +static const guid_t nouveau_op_dsm_muid = + GUID_INIT(0xA486D8F8, 0x0BDA, 0x471B, + 0xA7, 0x2B, 0x60, 0x42, 0xA6, 0xB5, 0xBE, 0xE0); static int nouveau_optimus_dsm(acpi_handle handle, int func, int arg, uint32_t *result) { @@ -86,7 +84,7 @@ static int nouveau_optimus_dsm(acpi_handle handle, int func, int arg, uint32_t * args_buff[i] = (arg >> i * 8) & 0xFF; *result = 0; - obj = acpi_evaluate_dsm_typed(handle, nouveau_op_dsm_muid, 0x00000100, + obj = acpi_evaluate_dsm_typed(handle, &nouveau_op_dsm_muid, 0x00000100, func, &argv4, ACPI_TYPE_BUFFER); if (!obj) { acpi_handle_info(handle, "failed to evaluate _DSM\n"); @@ -138,7 +136,7 @@ static int nouveau_dsm(acpi_handle handle, int func, int arg) .integer.value = arg, }; - obj = acpi_evaluate_dsm_typed(handle, nouveau_dsm_muid, 0x00000102, + obj = acpi_evaluate_dsm_typed(handle, &nouveau_dsm_muid, 0x00000102, func, &argv4, ACPI_TYPE_INTEGER); if (!obj) { acpi_handle_info(handle, "failed to evaluate _DSM\n"); @@ -259,7 +257,7 @@ static void nouveau_dsm_pci_probe(struct pci_dev *pdev, acpi_handle *dhandle_out if (!acpi_has_method(dhandle, "_DSM")) return; - supports_mux = acpi_check_dsm(dhandle, nouveau_dsm_muid, 0x00000102, + supports_mux = acpi_check_dsm(dhandle, &nouveau_dsm_muid, 0x00000102, 1 << NOUVEAU_DSM_POWER); optimus_funcs = nouveau_dsm_get_optimus_functions(dhandle); diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/mxm/base.c b/drivers/gpu/drm/nouveau/nvkm/subdev/mxm/base.c index e3e2f5e83815..f44682d62f75 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/mxm/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/mxm/base.c @@ -81,10 +81,9 @@ mxm_shadow_dsm(struct nvkm_mxm *mxm, u8 version) { struct nvkm_subdev *subdev = &mxm->subdev; struct nvkm_device *device = subdev->device; - static char muid[] = { - 0x00, 0xA4, 0x04, 0x40, 0x7D, 0x91, 0xF2, 0x4C, - 0xB8, 0x9C, 0x79, 0xB6, 0x2F, 0xD5, 0x56, 0x65 - }; + static guid_t muid = + GUID_INIT(0x4004A400, 0x917D, 0x4CF2, + 0xB8, 0x9C, 0x79, 0xB6, 0x2F, 0xD5, 0x56, 0x65); u32 mxms_args[] = { 0x00000000 }; union acpi_object argv4 = { .buffer.type = ACPI_TYPE_BUFFER, @@ -105,7 +104,7 @@ mxm_shadow_dsm(struct nvkm_mxm *mxm, u8 version) * unless you pass in exactly the version it supports.. */ rev = (version & 0xf0) << 4 | (version & 0x0f); - obj = acpi_evaluate_dsm(handle, muid, rev, 0x00000010, &argv4); + obj = acpi_evaluate_dsm(handle, &muid, rev, 0x00000010, &argv4); if (!obj) { nvkm_debug(subdev, "DSM MXMS failed\n"); return false; diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile index a5d3cd3ecb5f..4acbb944bcd2 100644 --- a/drivers/gpu/drm/radeon/Makefile +++ b/drivers/gpu/drm/radeon/Makefile @@ -105,7 +105,6 @@ radeon-y += \ vce_v2_0.o \ radeon_kfd.o -radeon-$(CONFIG_COMPAT) += radeon_ioc32.o radeon-$(CONFIG_VGA_SWITCHEROO) += radeon_atpx_handler.o radeon-$(CONFIG_ACPI) += radeon_acpi.o diff --git a/drivers/gpu/drm/radeon/atombios_encoders.c b/drivers/gpu/drm/radeon/atombios_encoders.c index fa4f8f008e4d..e67ed383e11b 100644 --- a/drivers/gpu/drm/radeon/atombios_encoders.c +++ b/drivers/gpu/drm/radeon/atombios_encoders.c @@ -31,6 +31,7 @@ #include "radeon_asic.h" #include "atom.h" #include <linux/backlight.h> +#include <linux/dmi.h> extern int atom_debug; @@ -2184,9 +2185,17 @@ int radeon_atom_pick_dig_encoder(struct drm_encoder *encoder, int fe_idx) goto assigned; } - /* on DCE32 and encoder can driver any block so just crtc id */ + /* + * On DCE32 any encoder can drive any block so usually just use crtc id, + * but Apple thinks different at least on iMac10,1, so there use linkb, + * otherwise the internal eDP panel will stay dark. + */ if (ASIC_IS_DCE32(rdev)) { - enc_idx = radeon_crtc->crtc_id; + if (dmi_match(DMI_PRODUCT_NAME, "iMac10,1")) + enc_idx = (dig->linkb) ? 1 : 0; + else + enc_idx = radeon_crtc->crtc_id; + goto assigned; } diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h index 68be1bfa22b9..5008f3d4cccc 100644 --- a/drivers/gpu/drm/radeon/radeon.h +++ b/drivers/gpu/drm/radeon/radeon.h @@ -377,7 +377,7 @@ struct radeon_fence { unsigned ring; bool is_vm_update; - wait_queue_t fence_wake; + wait_queue_entry_t fence_wake; }; int radeon_fence_driver_start_ring(struct radeon_device *rdev, int ring); diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c index 0a6444d72000..997131d58c7f 100644 --- a/drivers/gpu/drm/radeon/radeon_device.c +++ b/drivers/gpu/drm/radeon/radeon_device.c @@ -113,7 +113,6 @@ static inline bool radeon_is_atpx_hybrid(void) { return false; } #endif #define RADEON_PX_QUIRK_DISABLE_PX (1 << 0) -#define RADEON_PX_QUIRK_LONG_WAKEUP (1 << 1) struct radeon_px_quirk { u32 chip_vendor; @@ -140,8 +139,6 @@ static struct radeon_px_quirk radeon_px_quirk_list[] = { * https://bugs.freedesktop.org/show_bug.cgi?id=101491 */ { PCI_VENDOR_ID_ATI, 0x6741, 0x1043, 0x2122, RADEON_PX_QUIRK_DISABLE_PX }, - /* macbook pro 8.2 */ - { PCI_VENDOR_ID_ATI, 0x6741, PCI_VENDOR_ID_APPLE, 0x00e2, RADEON_PX_QUIRK_LONG_WAKEUP }, { 0, 0, 0, 0, 0 }, }; @@ -1245,25 +1242,17 @@ static void radeon_check_arguments(struct radeon_device *rdev) static void radeon_switcheroo_set_state(struct pci_dev *pdev, enum vga_switcheroo_state state) { struct drm_device *dev = pci_get_drvdata(pdev); - struct radeon_device *rdev = dev->dev_private; if (radeon_is_px(dev) && state == VGA_SWITCHEROO_OFF) return; if (state == VGA_SWITCHEROO_ON) { - unsigned d3_delay = dev->pdev->d3_delay; - pr_info("radeon: switched on\n"); /* don't suspend or resume card normally */ dev->switch_power_state = DRM_SWITCH_POWER_CHANGING; - if (d3_delay < 20 && (rdev->px_quirk_flags & RADEON_PX_QUIRK_LONG_WAKEUP)) - dev->pdev->d3_delay = 20; - radeon_resume_kms(dev, true, true); - dev->pdev->d3_delay = d3_delay; - dev->switch_power_state = DRM_SWITCH_POWER_ON; drm_kms_helper_poll_enable(dev); } else { diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c index ff2641cbf172..b401f1689bc1 100644 --- a/drivers/gpu/drm/radeon/radeon_drv.c +++ b/drivers/gpu/drm/radeon/radeon_drv.c @@ -38,6 +38,7 @@ #include <linux/module.h> #include <linux/pm_runtime.h> #include <linux/vga_switcheroo.h> +#include <linux/compat.h> #include <drm/drm_gem.h> #include <drm/drm_fb_helper.h> @@ -150,8 +151,6 @@ void radeon_gem_prime_unpin(struct drm_gem_object *obj); struct reservation_object *radeon_gem_prime_res_obj(struct drm_gem_object *); void *radeon_gem_prime_vmap(struct drm_gem_object *obj); void radeon_gem_prime_vunmap(struct drm_gem_object *obj, void *vaddr); -extern long radeon_kms_compat_ioctl(struct file *filp, unsigned int cmd, - unsigned long arg); /* atpx handler */ #if defined(CONFIG_VGA_SWITCHEROO) @@ -509,6 +508,21 @@ long radeon_drm_ioctl(struct file *filp, return ret; } +#ifdef CONFIG_COMPAT +static long radeon_kms_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) +{ + unsigned int nr = DRM_IOCTL_NR(cmd); + int ret; + + if (nr < DRM_COMMAND_BASE) + return drm_compat_ioctl(filp, cmd, arg); + + ret = radeon_drm_ioctl(filp, cmd, arg); + + return ret; +} +#endif + static const struct dev_pm_ops radeon_pm_ops = { .suspend = radeon_pmops_suspend, .resume = radeon_pmops_resume, diff --git a/drivers/gpu/drm/radeon/radeon_fence.c b/drivers/gpu/drm/radeon/radeon_fence.c index ef09f0a63754..e86f2bd38410 100644 --- a/drivers/gpu/drm/radeon/radeon_fence.c +++ b/drivers/gpu/drm/radeon/radeon_fence.c @@ -158,7 +158,7 @@ int radeon_fence_emit(struct radeon_device *rdev, * for the fence locking itself, so unlocked variants are used for * fence_signal, and remove_wait_queue. */ -static int radeon_fence_check_signaled(wait_queue_t *wait, unsigned mode, int flags, void *key) +static int radeon_fence_check_signaled(wait_queue_entry_t *wait, unsigned mode, int flags, void *key) { struct radeon_fence *fence; u64 seq; diff --git a/drivers/gpu/drm/radeon/radeon_ioc32.c b/drivers/gpu/drm/radeon/radeon_ioc32.c deleted file mode 100644 index 0b98ea134579..000000000000 --- a/drivers/gpu/drm/radeon/radeon_ioc32.c +++ /dev/null @@ -1,424 +0,0 @@ -/** - * \file radeon_ioc32.c - * - * 32-bit ioctl compatibility routines for the Radeon DRM. - * - * \author Paul Mackerras <paulus@samba.org> - * - * Copyright (C) Paul Mackerras 2005 - * All Rights Reserved. - * - * Permission is hereby granted, free of charge, to any person obtaining a - * copy of this software and associated documentation files (the "Software"), - * to deal in the Software without restriction, including without limitation - * the rights to use, copy, modify, merge, publish, distribute, sublicense, - * and/or sell copies of the Software, and to permit persons to whom the - * Software is furnished to do so, subject to the following conditions: - * - * The above copyright notice and this permission notice (including the next - * paragraph) shall be included in all copies or substantial portions of the - * Software. - * - * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR - * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, - * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL - * THE AUTHOR BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, - * WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, - * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS - * IN THE SOFTWARE. - */ -#include <linux/compat.h> - -#include <drm/drmP.h> -#include <drm/radeon_drm.h> -#include "radeon_drv.h" - -typedef struct drm_radeon_init32 { - int func; - u32 sarea_priv_offset; - int is_pci; - int cp_mode; - int gart_size; - int ring_size; - int usec_timeout; - - unsigned int fb_bpp; - unsigned int front_offset, front_pitch; - unsigned int back_offset, back_pitch; - unsigned int depth_bpp; - unsigned int depth_offset, depth_pitch; - - u32 fb_offset; - u32 mmio_offset; - u32 ring_offset; - u32 ring_rptr_offset; - u32 buffers_offset; - u32 gart_textures_offset; -} drm_radeon_init32_t; - -static int compat_radeon_cp_init(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_init32_t init32; - drm_radeon_init_t __user *init; - - if (copy_from_user(&init32, (void __user *)arg, sizeof(init32))) - return -EFAULT; - - init = compat_alloc_user_space(sizeof(*init)); - if (!access_ok(VERIFY_WRITE, init, sizeof(*init)) - || __put_user(init32.func, &init->func) - || __put_user(init32.sarea_priv_offset, &init->sarea_priv_offset) - || __put_user(init32.is_pci, &init->is_pci) - || __put_user(init32.cp_mode, &init->cp_mode) - || __put_user(init32.gart_size, &init->gart_size) - || __put_user(init32.ring_size, &init->ring_size) - || __put_user(init32.usec_timeout, &init->usec_timeout) - || __put_user(init32.fb_bpp, &init->fb_bpp) - || __put_user(init32.front_offset, &init->front_offset) - || __put_user(init32.front_pitch, &init->front_pitch) - || __put_user(init32.back_offset, &init->back_offset) - || __put_user(init32.back_pitch, &init->back_pitch) - || __put_user(init32.depth_bpp, &init->depth_bpp) - || __put_user(init32.depth_offset, &init->depth_offset) - || __put_user(init32.depth_pitch, &init->depth_pitch) - || __put_user(init32.fb_offset, &init->fb_offset) - || __put_user(init32.mmio_offset, &init->mmio_offset) - || __put_user(init32.ring_offset, &init->ring_offset) - || __put_user(init32.ring_rptr_offset, &init->ring_rptr_offset) - || __put_user(init32.buffers_offset, &init->buffers_offset) - || __put_user(init32.gart_textures_offset, - &init->gart_textures_offset)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_CP_INIT, (unsigned long)init); -} - -typedef struct drm_radeon_clear32 { - unsigned int flags; - unsigned int clear_color; - unsigned int clear_depth; - unsigned int color_mask; - unsigned int depth_mask; /* misnamed field: should be stencil */ - u32 depth_boxes; -} drm_radeon_clear32_t; - -static int compat_radeon_cp_clear(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_clear32_t clr32; - drm_radeon_clear_t __user *clr; - - if (copy_from_user(&clr32, (void __user *)arg, sizeof(clr32))) - return -EFAULT; - - clr = compat_alloc_user_space(sizeof(*clr)); - if (!access_ok(VERIFY_WRITE, clr, sizeof(*clr)) - || __put_user(clr32.flags, &clr->flags) - || __put_user(clr32.clear_color, &clr->clear_color) - || __put_user(clr32.clear_depth, &clr->clear_depth) - || __put_user(clr32.color_mask, &clr->color_mask) - || __put_user(clr32.depth_mask, &clr->depth_mask) - || __put_user((void __user *)(unsigned long)clr32.depth_boxes, - &clr->depth_boxes)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_CLEAR, (unsigned long)clr); -} - -typedef struct drm_radeon_stipple32 { - u32 mask; -} drm_radeon_stipple32_t; - -static int compat_radeon_cp_stipple(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_stipple32_t __user *argp = (void __user *)arg; - drm_radeon_stipple_t __user *request; - u32 mask; - - if (get_user(mask, &argp->mask)) - return -EFAULT; - - request = compat_alloc_user_space(sizeof(*request)); - if (!access_ok(VERIFY_WRITE, request, sizeof(*request)) - || __put_user((unsigned int __user *)(unsigned long)mask, - &request->mask)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_STIPPLE, (unsigned long)request); -} - -typedef struct drm_radeon_tex_image32 { - unsigned int x, y; /* Blit coordinates */ - unsigned int width, height; - u32 data; -} drm_radeon_tex_image32_t; - -typedef struct drm_radeon_texture32 { - unsigned int offset; - int pitch; - int format; - int width; /* Texture image coordinates */ - int height; - u32 image; -} drm_radeon_texture32_t; - -static int compat_radeon_cp_texture(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_texture32_t req32; - drm_radeon_texture_t __user *request; - drm_radeon_tex_image32_t img32; - drm_radeon_tex_image_t __user *image; - - if (copy_from_user(&req32, (void __user *)arg, sizeof(req32))) - return -EFAULT; - if (req32.image == 0) - return -EINVAL; - if (copy_from_user(&img32, (void __user *)(unsigned long)req32.image, - sizeof(img32))) - return -EFAULT; - - request = compat_alloc_user_space(sizeof(*request) + sizeof(*image)); - if (!access_ok(VERIFY_WRITE, request, - sizeof(*request) + sizeof(*image))) - return -EFAULT; - image = (drm_radeon_tex_image_t __user *) (request + 1); - - if (__put_user(req32.offset, &request->offset) - || __put_user(req32.pitch, &request->pitch) - || __put_user(req32.format, &request->format) - || __put_user(req32.width, &request->width) - || __put_user(req32.height, &request->height) - || __put_user(image, &request->image) - || __put_user(img32.x, &image->x) - || __put_user(img32.y, &image->y) - || __put_user(img32.width, &image->width) - || __put_user(img32.height, &image->height) - || __put_user((const void __user *)(unsigned long)img32.data, - &image->data)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_TEXTURE, (unsigned long)request); -} - -typedef struct drm_radeon_vertex2_32 { - int idx; /* Index of vertex buffer */ - int discard; /* Client finished with buffer? */ - int nr_states; - u32 state; - int nr_prims; - u32 prim; -} drm_radeon_vertex2_32_t; - -static int compat_radeon_cp_vertex2(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_vertex2_32_t req32; - drm_radeon_vertex2_t __user *request; - - if (copy_from_user(&req32, (void __user *)arg, sizeof(req32))) - return -EFAULT; - - request = compat_alloc_user_space(sizeof(*request)); - if (!access_ok(VERIFY_WRITE, request, sizeof(*request)) - || __put_user(req32.idx, &request->idx) - || __put_user(req32.discard, &request->discard) - || __put_user(req32.nr_states, &request->nr_states) - || __put_user((void __user *)(unsigned long)req32.state, - &request->state) - || __put_user(req32.nr_prims, &request->nr_prims) - || __put_user((void __user *)(unsigned long)req32.prim, - &request->prim)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_VERTEX2, (unsigned long)request); -} - -typedef struct drm_radeon_cmd_buffer32 { - int bufsz; - u32 buf; - int nbox; - u32 boxes; -} drm_radeon_cmd_buffer32_t; - -static int compat_radeon_cp_cmdbuf(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_cmd_buffer32_t req32; - drm_radeon_cmd_buffer_t __user *request; - - if (copy_from_user(&req32, (void __user *)arg, sizeof(req32))) - return -EFAULT; - - request = compat_alloc_user_space(sizeof(*request)); - if (!access_ok(VERIFY_WRITE, request, sizeof(*request)) - || __put_user(req32.bufsz, &request->bufsz) - || __put_user((void __user *)(unsigned long)req32.buf, - &request->buf) - || __put_user(req32.nbox, &request->nbox) - || __put_user((void __user *)(unsigned long)req32.boxes, - &request->boxes)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_CMDBUF, (unsigned long)request); -} - -typedef struct drm_radeon_getparam32 { - int param; - u32 value; -} drm_radeon_getparam32_t; - -static int compat_radeon_cp_getparam(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_getparam32_t req32; - drm_radeon_getparam_t __user *request; - - if (copy_from_user(&req32, (void __user *)arg, sizeof(req32))) - return -EFAULT; - - request = compat_alloc_user_space(sizeof(*request)); - if (!access_ok(VERIFY_WRITE, request, sizeof(*request)) - || __put_user(req32.param, &request->param) - || __put_user((void __user *)(unsigned long)req32.value, - &request->value)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_GETPARAM, (unsigned long)request); -} - -typedef struct drm_radeon_mem_alloc32 { - int region; - int alignment; - int size; - u32 region_offset; /* offset from start of fb or GART */ -} drm_radeon_mem_alloc32_t; - -static int compat_radeon_mem_alloc(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_mem_alloc32_t req32; - drm_radeon_mem_alloc_t __user *request; - - if (copy_from_user(&req32, (void __user *)arg, sizeof(req32))) - return -EFAULT; - - request = compat_alloc_user_space(sizeof(*request)); - if (!access_ok(VERIFY_WRITE, request, sizeof(*request)) - || __put_user(req32.region, &request->region) - || __put_user(req32.alignment, &request->alignment) - || __put_user(req32.size, &request->size) - || __put_user((int __user *)(unsigned long)req32.region_offset, - &request->region_offset)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_ALLOC, (unsigned long)request); -} - -typedef struct drm_radeon_irq_emit32 { - u32 irq_seq; -} drm_radeon_irq_emit32_t; - -static int compat_radeon_irq_emit(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_irq_emit32_t req32; - drm_radeon_irq_emit_t __user *request; - - if (copy_from_user(&req32, (void __user *)arg, sizeof(req32))) - return -EFAULT; - - request = compat_alloc_user_space(sizeof(*request)); - if (!access_ok(VERIFY_WRITE, request, sizeof(*request)) - || __put_user((int __user *)(unsigned long)req32.irq_seq, - &request->irq_seq)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_IRQ_EMIT, (unsigned long)request); -} - -/* The two 64-bit arches where alignof(u64)==4 in 32-bit code */ -#if defined (CONFIG_X86_64) || defined(CONFIG_IA64) -typedef struct drm_radeon_setparam32 { - int param; - u64 value; -} __attribute__((packed)) drm_radeon_setparam32_t; - -static int compat_radeon_cp_setparam(struct file *file, unsigned int cmd, - unsigned long arg) -{ - drm_radeon_setparam32_t req32; - drm_radeon_setparam_t __user *request; - - if (copy_from_user(&req32, (void __user *) arg, sizeof(req32))) - return -EFAULT; - - request = compat_alloc_user_space(sizeof(*request)); - if (!access_ok(VERIFY_WRITE, request, sizeof(*request)) - || __put_user(req32.param, &request->param) - || __put_user((void __user *)(unsigned long)req32.value, - &request->value)) - return -EFAULT; - - return drm_ioctl(file, DRM_IOCTL_RADEON_SETPARAM, (unsigned long) request); -} -#else -#define compat_radeon_cp_setparam NULL -#endif /* X86_64 || IA64 */ - -static drm_ioctl_compat_t *radeon_compat_ioctls[] = { - [DRM_RADEON_CP_INIT] = compat_radeon_cp_init, - [DRM_RADEON_CLEAR] = compat_radeon_cp_clear, - [DRM_RADEON_STIPPLE] = compat_radeon_cp_stipple, - [DRM_RADEON_TEXTURE] = compat_radeon_cp_texture, - [DRM_RADEON_VERTEX2] = compat_radeon_cp_vertex2, - [DRM_RADEON_CMDBUF] = compat_radeon_cp_cmdbuf, - [DRM_RADEON_GETPARAM] = compat_radeon_cp_getparam, - [DRM_RADEON_SETPARAM] = compat_radeon_cp_setparam, - [DRM_RADEON_ALLOC] = compat_radeon_mem_alloc, - [DRM_RADEON_IRQ_EMIT] = compat_radeon_irq_emit, -}; - -/** - * Called whenever a 32-bit process running under a 64-bit kernel - * performs an ioctl on /dev/dri/card<n>. - * - * \param filp file pointer. - * \param cmd command. - * \param arg user argument. - * \return zero on success or negative number on failure. - */ -long radeon_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) -{ - unsigned int nr = DRM_IOCTL_NR(cmd); - drm_ioctl_compat_t *fn = NULL; - int ret; - - if (nr < DRM_COMMAND_BASE) - return drm_compat_ioctl(filp, cmd, arg); - - if (nr < DRM_COMMAND_BASE + ARRAY_SIZE(radeon_compat_ioctls)) - fn = radeon_compat_ioctls[nr - DRM_COMMAND_BASE]; - - if (fn != NULL) - ret = (*fn) (filp, cmd, arg); - else - ret = drm_ioctl(filp, cmd, arg); - - return ret; -} - -long radeon_kms_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) -{ - unsigned int nr = DRM_IOCTL_NR(cmd); - int ret; - - if (nr < DRM_COMMAND_BASE) - return drm_compat_ioctl(filp, cmd, arg); - - ret = radeon_drm_ioctl(filp, cmd, arg); - - return ret; -} diff --git a/drivers/gpu/drm/rockchip/cdn-dp-core.c b/drivers/gpu/drm/rockchip/cdn-dp-core.c index 14fa1f8351e8..9b0b0588bbed 100644 --- a/drivers/gpu/drm/rockchip/cdn-dp-core.c +++ b/drivers/gpu/drm/rockchip/cdn-dp-core.c @@ -1195,7 +1195,7 @@ static int cdn_dp_probe(struct platform_device *pdev) continue; port = devm_kzalloc(dev, sizeof(*port), GFP_KERNEL); - if (!dp) + if (!port) return -ENOMEM; port->extcon = extcon; diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h index 47905faf5586..c7e96b82cf63 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_drv.h +++ b/drivers/gpu/drm/rockchip/rockchip_drm_drv.h @@ -45,13 +45,13 @@ struct rockchip_crtc_state { * * @crtc: array of enabled CRTCs, used to map from "pipe" to drm_crtc. * @num_pipe: number of pipes for this device. + * @mm_lock: protect drm_mm on multi-threads. */ struct rockchip_drm_private { struct drm_fb_helper fbdev_helper; struct drm_gem_object *fbdev_bo; struct drm_atomic_state *state; struct iommu_domain *domain; - /* protect drm_mm on multi-threads */ struct mutex mm_lock; struct drm_mm mm; struct list_head psr_list; diff --git a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c index df9e57064f19..b74ac717e56a 100644 --- a/drivers/gpu/drm/rockchip/rockchip_drm_gem.c +++ b/drivers/gpu/drm/rockchip/rockchip_drm_gem.c @@ -29,12 +29,11 @@ static int rockchip_gem_iommu_map(struct rockchip_gem_object *rk_obj) ssize_t ret; mutex_lock(&private->mm_lock); - ret = drm_mm_insert_node_generic(&private->mm, &rk_obj->mm, rk_obj->base.size, PAGE_SIZE, 0, 0); - mutex_unlock(&private->mm_lock); + if (ret < 0) { DRM_ERROR("out of I/O virtual memory: %zd\n", ret); return ret; @@ -56,7 +55,9 @@ static int rockchip_gem_iommu_map(struct rockchip_gem_object *rk_obj) return 0; err_remove_node: + mutex_lock(&private->mm_lock); drm_mm_remove_node(&rk_obj->mm); + mutex_unlock(&private->mm_lock); return ret; } diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c index a6d7fcb99c0b..22b57020790d 100644 --- a/drivers/gpu/drm/ttm/ttm_bo.c +++ b/drivers/gpu/drm/ttm/ttm_bo.c @@ -1353,7 +1353,6 @@ int ttm_bo_clean_mm(struct ttm_bo_device *bdev, unsigned mem_type) mem_type); return ret; } - dma_fence_put(man->move); man->use_type = false; man->has_type = false; @@ -1369,6 +1368,9 @@ int ttm_bo_clean_mm(struct ttm_bo_device *bdev, unsigned mem_type) ret = (*man->func->takedown)(man); } + dma_fence_put(man->move); + man->move = NULL; + return ret; } EXPORT_SYMBOL(ttm_bo_clean_mm); diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c index 13db8a2851ed..1f013d45c9e9 100644 --- a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c +++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf_res.c @@ -321,6 +321,7 @@ void vmw_cmdbuf_res_man_destroy(struct vmw_cmdbuf_res_manager *man) list_for_each_entry_safe(entry, next, &man->list, head) vmw_cmdbuf_res_free(man, entry); + drm_ht_remove(&man->resources); kfree(man); } |