diff options
Diffstat (limited to 'drivers/gpu/drm/vc4')
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_bo.c | 23 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_crtc.c | 19 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_drv.h | 50 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_gem.c | 176 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_hdmi.c | 42 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_irq.c | 60 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_regs.h | 18 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_render_cl.c | 22 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_v3d.c | 49 | ||||
| -rw-r--r-- | drivers/gpu/drm/vc4/vc4_validate.c | 4 |
10 files changed, 357 insertions, 106 deletions
diff --git a/drivers/gpu/drm/vc4/vc4_bo.c b/drivers/gpu/drm/vc4/vc4_bo.c index 18dfe3ec9a62..ac8eafea6361 100644 --- a/drivers/gpu/drm/vc4/vc4_bo.c +++ b/drivers/gpu/drm/vc4/vc4_bo.c @@ -215,7 +215,7 @@ struct vc4_bo *vc4_bo_create(struct drm_device *dev, size_t unaligned_size, struct drm_gem_cma_object *cma_obj; if (size == 0) - return NULL; + return ERR_PTR(-EINVAL); /* First, try to get a vc4_bo from the kernel BO cache. */ if (from_cache) { @@ -237,7 +237,7 @@ struct vc4_bo *vc4_bo_create(struct drm_device *dev, size_t unaligned_size, if (IS_ERR(cma_obj)) { DRM_ERROR("Failed to allocate from CMA:\n"); vc4_bo_stats_dump(vc4); - return NULL; + return ERR_PTR(-ENOMEM); } } @@ -259,8 +259,8 @@ int vc4_dumb_create(struct drm_file *file_priv, args->size = args->pitch * args->height; bo = vc4_bo_create(dev, args->size, false); - if (!bo) - return -ENOMEM; + if (IS_ERR(bo)) + return PTR_ERR(bo); ret = drm_gem_handle_create(file_priv, &bo->base.base, &args->handle); drm_gem_object_unreference_unlocked(&bo->base.base); @@ -443,8 +443,8 @@ int vc4_create_bo_ioctl(struct drm_device *dev, void *data, * get zeroed, and that might leak data between users. */ bo = vc4_bo_create(dev, args->size, false); - if (!bo) - return -ENOMEM; + if (IS_ERR(bo)) + return PTR_ERR(bo); ret = drm_gem_handle_create(file_priv, &bo->base.base, &args->handle); drm_gem_object_unreference_unlocked(&bo->base.base); @@ -496,14 +496,15 @@ vc4_create_shader_bo_ioctl(struct drm_device *dev, void *data, } bo = vc4_bo_create(dev, args->size, true); - if (!bo) - return -ENOMEM; + if (IS_ERR(bo)) + return PTR_ERR(bo); - ret = copy_from_user(bo->base.vaddr, + if (copy_from_user(bo->base.vaddr, (void __user *)(uintptr_t)args->data, - args->size); - if (ret != 0) + args->size)) { + ret = -EFAULT; goto fail; + } /* Clear the rest of the memory from allocating from the BO * cache. */ diff --git a/drivers/gpu/drm/vc4/vc4_crtc.c b/drivers/gpu/drm/vc4/vc4_crtc.c index 619dc781c517..355ee4b091b3 100644 --- a/drivers/gpu/drm/vc4/vc4_crtc.c +++ b/drivers/gpu/drm/vc4/vc4_crtc.c @@ -88,7 +88,7 @@ static const struct { } crtc_regs[] = { CRTC_REG(PV_CONTROL), CRTC_REG(PV_V_CONTROL), - CRTC_REG(PV_VSYNCD), + CRTC_REG(PV_VSYNCD_EVEN), CRTC_REG(PV_HORZA), CRTC_REG(PV_HORZB), CRTC_REG(PV_VERTA), @@ -188,6 +188,8 @@ static int vc4_get_clock_select(struct drm_crtc *crtc) static void vc4_crtc_mode_set_nofb(struct drm_crtc *crtc) { + struct drm_device *dev = crtc->dev; + struct vc4_dev *vc4 = to_vc4_dev(dev); struct vc4_crtc *vc4_crtc = to_vc4_crtc(crtc); struct drm_crtc_state *state = crtc->state; struct drm_display_mode *mode = &state->adjusted_mode; @@ -217,6 +219,16 @@ static void vc4_crtc_mode_set_nofb(struct drm_crtc *crtc) PV_HORZB_HFP) | VC4_SET_FIELD(mode->hdisplay, PV_HORZB_HACTIVE)); + CRTC_WRITE(PV_VERTA, + VC4_SET_FIELD(mode->vtotal - mode->vsync_end, + PV_VERTA_VBP) | + VC4_SET_FIELD(mode->vsync_end - mode->vsync_start, + PV_VERTA_VSYNC)); + CRTC_WRITE(PV_VERTB, + VC4_SET_FIELD(mode->vsync_start - mode->vdisplay, + PV_VERTB_VFP) | + VC4_SET_FIELD(vactive, PV_VERTB_VACTIVE)); + if (interlace) { CRTC_WRITE(PV_VERTA_EVEN, VC4_SET_FIELD(mode->vtotal - mode->vsync_end - 1, @@ -246,6 +258,10 @@ static void vc4_crtc_mode_set_nofb(struct drm_crtc *crtc) PV_CONTROL_FIFO_CLR | PV_CONTROL_EN); + HVS_WRITE(SCALER_DISPBKGNDX(vc4_crtc->channel), + SCALER_DISPBKGND_AUTOHS | + (interlace ? SCALER_DISPBKGND_INTERLACE : 0)); + if (debug_dump_regs) { DRM_INFO("CRTC %d regs after:\n", drm_crtc_index(crtc)); vc4_crtc_dump_regs(vc4_crtc); @@ -527,6 +543,7 @@ static int vc4_async_page_flip(struct drm_crtc *crtc, /* Make sure all other async modesetes have landed. */ ret = down_interruptible(&vc4->async_modeset); if (ret) { + drm_framebuffer_unreference(fb); kfree(flip_state); return ret; } diff --git a/drivers/gpu/drm/vc4/vc4_drv.h b/drivers/gpu/drm/vc4/vc4_drv.h index 3d1df6b1c4d3..fa2ad15d4f62 100644 --- a/drivers/gpu/drm/vc4/vc4_drv.h +++ b/drivers/gpu/drm/vc4/vc4_drv.h @@ -52,7 +52,7 @@ struct vc4_dev { /* Protects bo_cache and the BO stats. */ struct mutex bo_lock; - /* Sequence number for the last job queued in job_list. + /* Sequence number for the last job queued in bin_job_list. * Starts at 0 (no jobs emitted). */ uint64_t emit_seqno; @@ -62,11 +62,19 @@ struct vc4_dev { */ uint64_t finished_seqno; - /* List of all struct vc4_exec_info for jobs to be executed. - * The first job in the list is the one currently programmed - * into ct0ca/ct1ca for execution. + /* List of all struct vc4_exec_info for jobs to be executed in + * the binner. The first job in the list is the one currently + * programmed into ct0ca for execution. */ - struct list_head job_list; + struct list_head bin_job_list; + + /* List of all struct vc4_exec_info for jobs that have + * completed binning and are ready for rendering. The first + * job in the list is the one currently programmed into ct1ca + * for execution. + */ + struct list_head render_job_list; + /* List of the finished vc4_exec_infos waiting to be freed by * job_done_work. */ @@ -91,8 +99,12 @@ struct vc4_dev { struct vc4_bo *overflow_mem; struct work_struct overflow_mem_work; + int power_refcount; + + /* Mutex controlling the power refcount. */ + struct mutex power_lock; + struct { - uint32_t last_ct0ca, last_ct1ca; struct timer_list timer; struct work_struct reset_work; } hangcheck; @@ -142,6 +154,7 @@ struct vc4_seqno_cb { }; struct vc4_v3d { + struct vc4_dev *vc4; struct platform_device *pdev; void __iomem *regs; }; @@ -202,6 +215,11 @@ struct vc4_exec_info { /* Sequence number for this bin/render job. */ uint64_t seqno; + /* Last current addresses the hardware was processing when the + * hangcheck timer checked on us. + */ + uint32_t last_ct0ca, last_ct1ca; + /* Kernel-space copy of the ioctl arguments */ struct drm_vc4_submit_cl *args; @@ -286,11 +304,20 @@ struct vc4_exec_info { }; static inline struct vc4_exec_info * -vc4_first_job(struct vc4_dev *vc4) +vc4_first_bin_job(struct vc4_dev *vc4) +{ + if (list_empty(&vc4->bin_job_list)) + return NULL; + return list_first_entry(&vc4->bin_job_list, struct vc4_exec_info, head); +} + +static inline struct vc4_exec_info * +vc4_first_render_job(struct vc4_dev *vc4) { - if (list_empty(&vc4->job_list)) + if (list_empty(&vc4->render_job_list)) return NULL; - return list_first_entry(&vc4->job_list, struct vc4_exec_info, head); + return list_first_entry(&vc4->render_job_list, + struct vc4_exec_info, head); } /** @@ -404,7 +431,9 @@ int vc4_wait_seqno_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); int vc4_wait_bo_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); -void vc4_submit_next_job(struct drm_device *dev); +void vc4_submit_next_bin_job(struct drm_device *dev); +void vc4_submit_next_render_job(struct drm_device *dev); +void vc4_move_job_to_render(struct drm_device *dev, struct vc4_exec_info *exec); int vc4_wait_for_seqno(struct drm_device *dev, uint64_t seqno, uint64_t timeout_ns, bool interruptible); void vc4_job_handle_completed(struct vc4_dev *vc4); @@ -443,7 +472,6 @@ void vc4_plane_async_set_fb(struct drm_plane *plane, extern struct platform_driver vc4_v3d_driver; int vc4_v3d_debugfs_ident(struct seq_file *m, void *unused); int vc4_v3d_debugfs_regs(struct seq_file *m, void *unused); -int vc4_v3d_set_power(struct vc4_dev *vc4, bool on); /* vc4_validate.c */ int diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c index 48ce30a6f4b5..8d4384f8b78d 100644 --- a/drivers/gpu/drm/vc4/vc4_gem.c +++ b/drivers/gpu/drm/vc4/vc4_gem.c @@ -23,6 +23,7 @@ #include <linux/module.h> #include <linux/platform_device.h> +#include <linux/pm_runtime.h> #include <linux/device.h> #include <linux/io.h> @@ -140,10 +141,10 @@ vc4_save_hang_state(struct drm_device *dev) struct vc4_dev *vc4 = to_vc4_dev(dev); struct drm_vc4_get_hang_state *state; struct vc4_hang_state *kernel_state; - struct vc4_exec_info *exec; + struct vc4_exec_info *exec[2]; struct vc4_bo *bo; unsigned long irqflags; - unsigned int i, unref_list_count; + unsigned int i, j, unref_list_count, prev_idx; kernel_state = kcalloc(1, sizeof(*kernel_state), GFP_KERNEL); if (!kernel_state) @@ -152,37 +153,55 @@ vc4_save_hang_state(struct drm_device *dev) state = &kernel_state->user_state; spin_lock_irqsave(&vc4->job_lock, irqflags); - exec = vc4_first_job(vc4); - if (!exec) { + exec[0] = vc4_first_bin_job(vc4); + exec[1] = vc4_first_render_job(vc4); + if (!exec[0] && !exec[1]) { spin_unlock_irqrestore(&vc4->job_lock, irqflags); return; } - unref_list_count = 0; - list_for_each_entry(bo, &exec->unref_list, unref_head) - unref_list_count++; + /* Get the bos from both binner and renderer into hang state. */ + state->bo_count = 0; + for (i = 0; i < 2; i++) { + if (!exec[i]) + continue; + + unref_list_count = 0; + list_for_each_entry(bo, &exec[i]->unref_list, unref_head) + unref_list_count++; + state->bo_count += exec[i]->bo_count + unref_list_count; + } + + kernel_state->bo = kcalloc(state->bo_count, + sizeof(*kernel_state->bo), GFP_ATOMIC); - state->bo_count = exec->bo_count + unref_list_count; - kernel_state->bo = kcalloc(state->bo_count, sizeof(*kernel_state->bo), - GFP_ATOMIC); if (!kernel_state->bo) { spin_unlock_irqrestore(&vc4->job_lock, irqflags); return; } - for (i = 0; i < exec->bo_count; i++) { - drm_gem_object_reference(&exec->bo[i]->base); - kernel_state->bo[i] = &exec->bo[i]->base; - } + prev_idx = 0; + for (i = 0; i < 2; i++) { + if (!exec[i]) + continue; - list_for_each_entry(bo, &exec->unref_list, unref_head) { - drm_gem_object_reference(&bo->base.base); - kernel_state->bo[i] = &bo->base.base; - i++; + for (j = 0; j < exec[i]->bo_count; j++) { + drm_gem_object_reference(&exec[i]->bo[j]->base); + kernel_state->bo[j + prev_idx] = &exec[i]->bo[j]->base; + } + + list_for_each_entry(bo, &exec[i]->unref_list, unref_head) { + drm_gem_object_reference(&bo->base.base); + kernel_state->bo[j + prev_idx] = &bo->base.base; + j++; + } + prev_idx = j + 1; } - state->start_bin = exec->ct0ca; - state->start_render = exec->ct1ca; + if (exec[0]) + state->start_bin = exec[0]->ct0ca; + if (exec[1]) + state->start_render = exec[1]->ct1ca; spin_unlock_irqrestore(&vc4->job_lock, irqflags); @@ -228,8 +247,16 @@ vc4_reset(struct drm_device *dev) struct vc4_dev *vc4 = to_vc4_dev(dev); DRM_INFO("Resetting GPU.\n"); - vc4_v3d_set_power(vc4, false); - vc4_v3d_set_power(vc4, true); + + mutex_lock(&vc4->power_lock); + if (vc4->power_refcount) { + /* Power the device off and back on the by dropping the + * reference on runtime PM. + */ + pm_runtime_put_sync_suspend(&vc4->v3d->pdev->dev); + pm_runtime_get_sync(&vc4->v3d->pdev->dev); + } + mutex_unlock(&vc4->power_lock); vc4_irq_reset(dev); @@ -257,10 +284,19 @@ vc4_hangcheck_elapsed(unsigned long data) struct drm_device *dev = (struct drm_device *)data; struct vc4_dev *vc4 = to_vc4_dev(dev); uint32_t ct0ca, ct1ca; + unsigned long irqflags; + struct vc4_exec_info *bin_exec, *render_exec; + + spin_lock_irqsave(&vc4->job_lock, irqflags); + + bin_exec = vc4_first_bin_job(vc4); + render_exec = vc4_first_render_job(vc4); /* If idle, we can stop watching for hangs. */ - if (list_empty(&vc4->job_list)) + if (!bin_exec && !render_exec) { + spin_unlock_irqrestore(&vc4->job_lock, irqflags); return; + } ct0ca = V3D_READ(V3D_CTNCA(0)); ct1ca = V3D_READ(V3D_CTNCA(1)); @@ -268,14 +304,19 @@ vc4_hangcheck_elapsed(unsigned long data) /* If we've made any progress in execution, rearm the timer * and wait. */ - if (ct0ca != vc4->hangcheck.last_ct0ca || - ct1ca != vc4->hangcheck.last_ct1ca) { - vc4->hangcheck.last_ct0ca = ct0ca; - vc4->hangcheck.last_ct1ca = ct1ca; + if ((bin_exec && ct0ca != bin_exec->last_ct0ca) || + (render_exec && ct1ca != render_exec->last_ct1ca)) { + if (bin_exec) + bin_exec->last_ct0ca = ct0ca; + if (render_exec) + render_exec->last_ct1ca = ct1ca; + spin_unlock_irqrestore(&vc4->job_lock, irqflags); vc4_queue_hangcheck(dev); return; } + spin_unlock_irqrestore(&vc4->job_lock, irqflags); + /* We've gone too long with no progress, reset. This has to * be done from a work struct, since resetting can sleep and * this timer hook isn't allowed to. @@ -340,12 +381,7 @@ vc4_wait_for_seqno(struct drm_device *dev, uint64_t seqno, uint64_t timeout_ns, finish_wait(&vc4->job_wait_queue, &wait); trace_vc4_wait_for_seqno_end(dev, seqno); - if (ret && ret != -ERESTARTSYS) { - DRM_ERROR("timeout waiting for render thread idle\n"); - return ret; - } - - return 0; + return ret; } static void @@ -373,11 +409,13 @@ vc4_flush_caches(struct drm_device *dev) * The job_lock should be held during this. */ void -vc4_submit_next_job(struct drm_device *dev) +vc4_submit_next_bin_job(struct drm_device *dev) { struct vc4_dev *vc4 = to_vc4_dev(dev); - struct vc4_exec_info *exec = vc4_first_job(vc4); + struct vc4_exec_info *exec; +again: + exec = vc4_first_bin_job(vc4); if (!exec) return; @@ -387,11 +425,40 @@ vc4_submit_next_job(struct drm_device *dev) V3D_WRITE(V3D_BPOA, 0); V3D_WRITE(V3D_BPOS, 0); - if (exec->ct0ca != exec->ct0ea) + /* Either put the job in the binner if it uses the binner, or + * immediately move it to the to-be-rendered queue. + */ + if (exec->ct0ca != exec->ct0ea) { submit_cl(dev, 0, exec->ct0ca, exec->ct0ea); + } else { + vc4_move_job_to_render(dev, exec); + goto again; + } +} + +void +vc4_submit_next_render_job(struct drm_device *dev) +{ + struct vc4_dev *vc4 = to_vc4_dev(dev); + struct vc4_exec_info *exec = vc4_first_render_job(vc4); + + if (!exec) + return; + submit_cl(dev, 1, exec->ct1ca, exec->ct1ea); } +void +vc4_move_job_to_render(struct drm_device *dev, struct vc4_exec_info *exec) +{ + struct vc4_dev *vc4 = to_vc4_dev(dev); + bool was_empty = list_empty(&vc4->render_job_list); + + list_move_tail(&exec->head, &vc4->render_job_list); + if (was_empty) + vc4_submit_next_render_job(dev); +} + static void vc4_update_bo_seqnos(struct vc4_exec_info *exec, uint64_t seqno) { @@ -430,14 +497,14 @@ vc4_queue_submit(struct drm_device *dev, struct vc4_exec_info *exec) exec->seqno = seqno; vc4_update_bo_seqnos(exec, seqno); - list_add_tail(&exec->head, &vc4->job_list); + list_add_tail(&exec->head, &vc4->bin_job_list); /* If no job was executing, kick ours off. Otherwise, it'll - * get started when the previous job's frame done interrupt + * get started when the previous job's flush done interrupt * occurs. */ - if (vc4_first_job(vc4) == exec) { - vc4_submit_next_job(dev); + if (vc4_first_bin_job(vc4) == exec) { + vc4_submit_next_bin_job(dev); vc4_queue_hangcheck(dev); } @@ -578,9 +645,9 @@ vc4_get_bcl(struct drm_device *dev, struct vc4_exec_info *exec) } bo = vc4_bo_create(dev, exec_size, true); - if (!bo) { + if (IS_ERR(bo)) { DRM_ERROR("Couldn't allocate BO for binning\n"); - ret = -ENOMEM; + ret = PTR_ERR(bo); goto fail; } exec->exec_bo = &bo->base; @@ -617,6 +684,7 @@ fail: static void vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec) { + struct vc4_dev *vc4 = to_vc4_dev(dev); unsigned i; /* Need the struct lock for drm_gem_object_unreference(). */ @@ -635,6 +703,11 @@ vc4_complete_exec(struct drm_device *dev, struct vc4_exec_info *exec) } mutex_unlock(&dev->struct_mutex); + mutex_lock(&vc4->power_lock); + if (--vc4->power_refcount == 0) + pm_runtime_put(&vc4->v3d->pdev->dev); + mutex_unlock(&vc4->power_lock); + kfree(exec); } @@ -746,6 +819,9 @@ vc4_wait_bo_ioctl(struct drm_device *dev, void *data, struct drm_gem_object *gem_obj; struct vc4_bo *bo; + if (args->pad != 0) + return -EINVAL; + gem_obj = drm_gem_object_lookup(dev, file_priv, args->handle); if (!gem_obj) { DRM_ERROR("Failed to look up GEM BO %d\n", args->handle); @@ -772,7 +848,7 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data, struct vc4_dev *vc4 = to_vc4_dev(dev); struct drm_vc4_submit_cl *args = data; struct vc4_exec_info *exec; - int ret; + int ret = 0; if ((args->flags & ~VC4_SUBMIT_CL_USE_CLEAR_COLOR) != 0) { DRM_ERROR("Unknown flags: 0x%02x\n", args->flags); @@ -785,6 +861,15 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data, return -ENOMEM; } + mutex_lock(&vc4->power_lock); + if (vc4->power_refcount++ == 0) + ret = pm_runtime_get_sync(&vc4->v3d->pdev->dev); + mutex_unlock(&vc4->power_lock); + if (ret < 0) { + kfree(exec); + return ret; + } + exec->args = args; INIT_LIST_HEAD(&exec->unref_list); @@ -828,7 +913,8 @@ vc4_gem_init(struct drm_device *dev) { struct vc4_dev *vc4 = to_vc4_dev(dev); - INIT_LIST_HEAD(&vc4->job_list); + INIT_LIST_HEAD(&vc4->bin_job_list); + INIT_LIST_HEAD(&vc4->render_job_list); INIT_LIST_HEAD(&vc4->job_done_list); INIT_LIST_HEAD(&vc4->seqno_cb_list); spin_lock_init(&vc4->job_lock); @@ -839,6 +925,8 @@ vc4_gem_init(struct drm_device *dev) (unsigned long)dev); INIT_WORK(&vc4->job_done_work, vc4_job_done_work); + + mutex_init(&vc4->power_lock); } void diff --git a/drivers/gpu/drm/vc4/vc4_hdmi.c b/drivers/gpu/drm/vc4/vc4_hdmi.c index c69c0460196b..d8b864925fd3 100644 --- a/drivers/gpu/drm/vc4/vc4_hdmi.c +++ b/drivers/gpu/drm/vc4/vc4_hdmi.c @@ -47,6 +47,7 @@ struct vc4_hdmi { void __iomem *hdmicore_regs; void __iomem *hd_regs; int hpd_gpio; + bool hpd_active_low; struct clk *pixel_clock; struct clk *hsm_clock; @@ -95,6 +96,7 @@ static const struct { HDMI_REG(VC4_HDMI_SW_RESET_CONTROL), HDMI_REG(VC4_HDMI_HOTPLUG_INT), HDMI_REG(VC4_HDMI_HOTPLUG), + HDMI_REG(VC4_HDMI_RAM_PACKET_CONFIG), HDMI_REG(VC4_HDMI_HORZA), HDMI_REG(VC4_HDMI_HORZB), HDMI_REG(VC4_HDMI_FIFO_CTL), @@ -165,7 +167,8 @@ vc4_hdmi_connector_detect(struct drm_connector *connector, bool force) struct vc4_dev *vc4 = to_vc4_dev(dev); if (vc4->hdmi->hpd_gpio) { - if (gpio_get_value(vc4->hdmi->hpd_gpio)) + if (gpio_get_value_cansleep(vc4->hdmi->hpd_gpio) ^ + vc4->hdmi->hpd_active_low) return connector_status_connected; else return connector_status_disconnected; @@ -495,6 +498,16 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data) goto err_put_i2c; } + /* This is the rate that is set by the firmware. The number + * needs to be a bit higher than the pixel clock rate + * (generally 148.5Mhz). + */ + ret = clk_set_rate(hdmi->hsm_clock, 163682864); + if (ret) { + DRM_ERROR("Failed to set HSM clock rate: %d\n", ret); + goto err_unprepare_pix; + } + ret = clk_prepare_enable(hdmi->hsm_clock); if (ret) { DRM_ERROR("Failed to turn on HDMI state machine clock: %d\n", @@ -506,17 +519,40 @@ static int vc4_hdmi_bind(struct device *dev, struct device *master, void *data) * we'll use the HDMI core's register. */ if (of_find_property(dev->of_node, "hpd-gpios", &value)) { - hdmi->hpd_gpio = of_get_named_gpio(dev->of_node, "hpd-gpios", 0); + enum of_gpio_flags hpd_gpio_flags; + + hdmi->hpd_gpio = of_get_named_gpio_flags(dev->of_node, + "hpd-gpios", 0, + &hpd_gpio_flags); if (hdmi->hpd_gpio < 0) { ret = hdmi->hpd_gpio; goto err_unprepare_hsm; } + + hdmi->hpd_active_low = hpd_gpio_flags & OF_GPIO_ACTIVE_LOW; } vc4->hdmi = hdmi; /* HDMI core must be enabled. */ - WARN_ON_ONCE((HD_READ(VC4_HD_M_CTL) & VC4_HD_M_ENABLE) == 0); + if (!(HD_READ(VC4_HD_M_CTL) & VC4_HD_M_ENABLE)) { + HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_SW_RST); + udelay(1); + HD_WRITE(VC4_HD_M_CTL, 0); + + HD_WRITE(VC4_HD_M_CTL, VC4_HD_M_ENABLE); + + HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL, + VC4_HDMI_SW_RESET_HDMI | + VC4_HDMI_SW_RESET_FORMAT_DETECT); + + HDMI_WRITE(VC4_HDMI_SW_RESET_CONTROL, 0); + + /* PHY should be in reset, like + * vc4_hdmi_encoder_disable() does. + */ + HDMI_WRITE(VC4_HDMI_TX_PHY_RESET_CTL, 0xf << 16); + } drm_encoder_init(drm, hdmi->encoder, &vc4_hdmi_encoder_funcs, DRM_MODE_ENCODER_TMDS, NULL); diff --git a/drivers/gpu/drm/vc4/vc4_irq.c b/drivers/gpu/drm/vc4/vc4_irq.c index b68060e758db..b0104a346a74 100644 --- a/drivers/gpu/drm/vc4/vc4_irq.c +++ b/drivers/gpu/drm/vc4/vc4_irq.c @@ -30,6 +30,10 @@ * disables that specific interrupt, and 0s written are ignored * (reading either one returns the set of enabled interrupts). * + * When we take a binning flush done interrupt, we need to submit the + * next frame for binning and move the finished frame to the render + * thread. + * * When we take a render frame interrupt, we need to wake the * processes waiting for some frame to be done, and get the next frame * submitted ASAP (so the hardware doesn't sit idle when there's work @@ -44,6 +48,7 @@ #include "vc4_regs.h" #define V3D_DRIVER_IRQS (V3D_INT_OUTOMEM | \ + V3D_INT_FLDONE | \ V3D_INT_FRDONE) DECLARE_WAIT_QUEUE_HEAD(render_wait); @@ -57,7 +62,7 @@ vc4_overflow_mem_work(struct work_struct *work) struct vc4_bo *bo; bo = vc4_bo_create(dev, 256 * 1024, true); - if (!bo) { + if (IS_ERR(bo)) { DRM_ERROR("Couldn't allocate binner overflow mem\n"); return; } @@ -77,7 +82,7 @@ vc4_overflow_mem_work(struct work_struct *work) unsigned long irqflags; spin_lock_irqsave(&vc4->job_lock, irqflags); - current_exec = vc4_first_job(vc4); + current_exec = vc4_first_bin_job(vc4); if (current_exec) { vc4->overflow_mem->seqno = vc4->finished_seqno + 1; list_add_tail(&vc4->overflow_mem->unref_head, @@ -98,17 +103,43 @@ vc4_overflow_mem_work(struct work_struct *work) } static void -vc4_irq_finish_job(struct drm_device *dev) +vc4_irq_finish_bin_job(struct drm_device *dev) +{ + struct vc4_dev *vc4 = to_vc4_dev(dev); + struct vc4_exec_info *exec = vc4_first_bin_job(vc4); + + if (!exec) + return; + + vc4_move_job_to_render(dev, exec); + vc4_submit_next_bin_job(dev); +} + +static void +vc4_cancel_bin_job(struct drm_device *dev) +{ + struct vc4_dev *vc4 = to_vc4_dev(dev); + struct vc4_exec_info *exec = vc4_first_bin_job(vc4); + + if (!exec) + return; + + list_move_tail(&exec->head, &vc4->bin_job_list); + vc4_submit_next_bin_job(dev); +} + +static void +vc4_irq_finish_render_job(struct drm_device *dev) { struct vc4_dev *vc4 = to_vc4_dev(dev); - struct vc4_exec_info *exec = vc4_first_job(vc4); + struct vc4_exec_info *exec = vc4_first_render_job(vc4); if (!exec) return; vc4->finished_seqno++; list_move_tail(&exec->head, &vc4->job_done_list); - vc4_submit_next_job(dev); + vc4_submit_next_render_job(dev); wake_up_all(&vc4->job_wait_queue); schedule_work(&vc4->job_done_work); @@ -125,9 +156,10 @@ vc4_irq(int irq, void *arg) barrier(); intctl = V3D_READ(V3D_INTCTL); - /* Acknowledge the interrupts we're handling here. The render - * frame done interrupt will be cleared, while OUTOMEM will - * stay high until the underlying cause is cleared. + /* Acknowledge the interrupts we're handling here. The binner + * last flush / render frame done interrupt will be cleared, + * while OUTOMEM will stay high until the underlying cause is + * cleared. */ V3D_WRITE(V3D_INTCTL, intctl); @@ -138,9 +170,16 @@ vc4_irq(int irq, void *arg) status = IRQ_HANDLED; } + if (intctl & V3D_INT_FLDONE) { + spin_lock(&vc4->job_lock); + vc4_irq_finish_bin_job(dev); + spin_unlock(&vc4->job_lock); + status = IRQ_HANDLED; + } + if (intctl & V3D_INT_FRDONE) { spin_lock(&vc4->job_lock); - vc4_irq_finish_job(dev); + vc4_irq_finish_render_job(dev); spin_unlock(&vc4->job_lock); status = IRQ_HANDLED; } @@ -205,6 +244,7 @@ void vc4_irq_reset(struct drm_device *dev) V3D_WRITE(V3D_INTENA, V3D_DRIVER_IRQS); spin_lock_irqsave(&vc4->job_lock, irqflags); - vc4_irq_finish_job(dev); + vc4_cancel_bin_job(dev); + vc4_irq_finish_render_job(dev); spin_unlock_irqrestore(&vc4->job_lock, irqflags); } diff --git a/drivers/gpu/drm/vc4/vc4_regs.h b/drivers/gpu/drm/vc4/vc4_regs.h index 25df20ef939c..bf42a8e87111 100644 --- a/drivers/gpu/drm/vc4/vc4_regs.h +++ b/drivers/gpu/drm/vc4/vc4_regs.h @@ -187,7 +187,7 @@ # define PV_VCONTROL_CONTINUOUS BIT(1) # define PV_VCONTROL_VIDEN BIT(0) -#define PV_VSYNCD 0x08 +#define PV_VSYNCD_EVEN 0x08 #define PV_HORZA 0x0c # define PV_HORZA_HBP_MASK VC4_MASK(31, 16) @@ -350,6 +350,17 @@ # define SCALER_DISPCTRLX_HEIGHT_SHIFT 0 #define SCALER_DISPBKGND0 0x00000044 +# define SCALER_DISPBKGND_AUTOHS BIT(31) +# define SCALER_DISPBKGND_INTERLACE BIT(30) +# define SCALER_DISPBKGND_GAMMA BIT(29) +# define SCALER_DISPBKGND_TESTMODE_MASK VC4_MASK(28, 25) +# define SCALER_DISPBKGND_TESTMODE_SHIFT 25 +/* Enables filling the scaler line with the RGB value in the low 24 + * bits before compositing. Costs cycles, so should be skipped if + * opaque display planes will cover everything. + */ +# define SCALER_DISPBKGND_FILL BIT(24) + #define SCALER_DISPSTAT0 0x00000048 #define SCALER_DISPBASE0 0x0000004c # define SCALER_DISPSTATX_MODE_MASK VC4_MASK(31, 30) @@ -362,6 +373,9 @@ # define SCALER_DISPSTATX_EMPTY BIT(28) #define SCALER_DISPCTRL1 0x00000050 #define SCALER_DISPBKGND1 0x00000054 +#define SCALER_DISPBKGNDX(x) (SCALER_DISPBKGND0 + \ + (x) * (SCALER_DISPBKGND1 - \ + SCALER_DISPBKGND0)) #define SCALER_DISPSTAT1 0x00000058 #define SCALER_DISPSTATX(x) (SCALER_DISPSTAT0 + \ (x) * (SCALER_DISPSTAT1 - \ @@ -456,6 +470,8 @@ #define VC4_HDMI_TX_PHY_RESET_CTL 0x2c0 #define VC4_HD_M_CTL 0x00c +# define VC4_HD_M_REGISTER_FILE_STANDBY (3 << 6) +# define VC4_HD_M_RAM_STANDBY (3 << 4) # define VC4_HD_M_SW_RST BIT(2) # define VC4_HD_M_ENABLE BIT(0) diff --git a/drivers/gpu/drm/vc4/vc4_render_cl.c b/drivers/gpu/drm/vc4/vc4_render_cl.c index 8a2a312e2c1b..0f12418725e5 100644 --- a/drivers/gpu/drm/vc4/vc4_render_cl.c +++ b/drivers/gpu/drm/vc4/vc4_render_cl.c @@ -316,20 +316,11 @@ static int vc4_create_rcl_bo(struct drm_device *dev, struct vc4_exec_info *exec, size += xtiles * ytiles * loop_body_size; setup->rcl = &vc4_bo_create(dev, size, true)->base; - if (!setup->rcl) - return -ENOMEM; + if (IS_ERR(setup->rcl)) + return PTR_ERR(setup->rcl); list_add_tail(&to_vc4_bo(&setup->rcl->base)->unref_head, &exec->unref_list); - rcl_u8(setup, VC4_PACKET_TILE_RENDERING_MODE_CONFIG); - rcl_u32(setup, - (setup->color_write ? (setup->color_write->paddr + - args->color_write.offset) : - 0)); - rcl_u16(setup, args->width); - rcl_u16(setup, args->height); - rcl_u16(setup, args->color_write.bits); - /* The tile buffer gets cleared when the previous tile is stored. If * the clear values changed between frames, then the tile buffer has * stale clear values in it, so we have to do a store in None mode (no @@ -349,6 +340,15 @@ static int vc4_create_rcl_bo(struct drm_device *dev, struct vc4_exec_info *exec, rcl_u32(setup, 0); /* no address, since we're in None mode */ } + rcl_u8(setup, VC4_PACKET_TILE_RENDERING_MODE_CONFIG); + rcl_u32(setup, + (setup->color_write ? (setup->color_write->paddr + + args->color_write.offset) : + 0)); + rcl_u16(setup, args->width); + rcl_u16(setup, args->height); + rcl_u16(setup, args->color_write.bits); + for (y = min_y_tile; y <= max_y_tile; y++) { for (x = min_x_tile; x <= max_x_tile; x++) { bool first = (x == min_x_tile && y == min_y_tile); diff --git a/drivers/gpu/drm/vc4/vc4_v3d.c b/drivers/gpu/drm/vc4/vc4_v3d.c index 314ff71db978..e6d3c6028341 100644 --- a/drivers/gpu/drm/vc4/vc4_v3d.c +++ b/drivers/gpu/drm/vc4/vc4_v3d.c @@ -17,6 +17,7 @@ */ #include "linux/component.h" +#include "linux/pm_runtime.h" #include "vc4_drv.h" #include "vc4_regs.h" @@ -144,18 +145,6 @@ int vc4_v3d_debugfs_ident(struct seq_file *m, void *unused) } #endif /* CONFIG_DEBUG_FS */ -int -vc4_v3d_set_power(struct vc4_dev *vc4, bool on) -{ - /* XXX: This interface is needed for GPU reset, and the way to - * do it is to turn our power domain off and back on. We - * can't just reset from within the driver, because the reset - * bits are in the power domain's register area, and get set - * during the poweron process. - */ - return 0; -} - static void vc4_v3d_init_hw(struct drm_device *dev) { struct vc4_dev *vc4 = to_vc4_dev(dev); @@ -167,6 +156,29 @@ static void vc4_v3d_init_hw(struct drm_device *dev) V3D_WRITE(V3D_VPMBASE, 0); } +#ifdef CONFIG_PM +static int vc4_v3d_runtime_suspend(struct device *dev) +{ + struct vc4_v3d *v3d = dev_get_drvdata(dev); + struct vc4_dev *vc4 = v3d->vc4; + + vc4_irq_uninstall(vc4->dev); + + return 0; +} + +static int vc4_v3d_runtime_resume(struct device *dev) +{ + struct vc4_v3d *v3d = dev_get_drvdata(dev); + struct vc4_dev *vc4 = v3d->vc4; + + vc4_v3d_init_hw(vc4->dev); + vc4_irq_postinstall(vc4->dev); + + return 0; +} +#endif + static int vc4_v3d_bind(struct device *dev, struct device *master, void *data) { struct platform_device *pdev = to_platform_device(dev); @@ -179,6 +191,8 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data) if (!v3d) return -ENOMEM; + dev_set_drvdata(dev, v3d); + v3d->pdev = pdev; v3d->regs = vc4_ioremap_regs(pdev, 0); @@ -186,6 +200,7 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data) return PTR_ERR(v3d->regs); vc4->v3d = v3d; + v3d->vc4 = vc4; if (V3D_READ(V3D_IDENT0) != V3D_EXPECTED_IDENT0) { DRM_ERROR("V3D_IDENT0 read 0x%08x instead of 0x%08x\n", @@ -207,6 +222,8 @@ static int vc4_v3d_bind(struct device *dev, struct device *master, void *data) return ret; } + pm_runtime_enable(dev); + return 0; } @@ -216,6 +233,8 @@ static void vc4_v3d_unbind(struct device *dev, struct device *master, struct drm_device *drm = dev_get_drvdata(master); struct vc4_dev *vc4 = to_vc4_dev(drm); + pm_runtime_disable(dev); + drm_irq_uninstall(drm); /* Disable the binner's overflow memory address, so the next @@ -228,6 +247,10 @@ static void vc4_v3d_unbind(struct device *dev, struct device *master, vc4->v3d = NULL; } +static const struct dev_pm_ops vc4_v3d_pm_ops = { + SET_RUNTIME_PM_OPS(vc4_v3d_runtime_suspend, vc4_v3d_runtime_resume, NULL) +}; + static const struct component_ops vc4_v3d_ops = { .bind = vc4_v3d_bind, .unbind = vc4_v3d_unbind, @@ -245,6 +268,7 @@ static int vc4_v3d_dev_remove(struct platform_device *pdev) } static const struct of_device_id vc4_v3d_dt_match[] = { + { .compatible = "brcm,bcm2835-v3d" }, { .compatible = "brcm,vc4-v3d" }, {} }; @@ -255,5 +279,6 @@ struct platform_driver vc4_v3d_driver = { .driver = { .name = "vc4_v3d", .of_match_table = vc4_v3d_dt_match, + .pm = &vc4_v3d_pm_ops, }, }; diff --git a/drivers/gpu/drm/vc4/vc4_validate.c b/drivers/gpu/drm/vc4/vc4_validate.c index e26d9f6face3..24c2c746e8f3 100644 --- a/drivers/gpu/drm/vc4/vc4_validate.c +++ b/drivers/gpu/drm/vc4/vc4_validate.c @@ -401,8 +401,8 @@ validate_tile_binning_config(VALIDATE_ARGS) tile_bo = vc4_bo_create(dev, exec->tile_alloc_offset + tile_alloc_size, true); exec->tile_bo = &tile_bo->base; - if (!exec->tile_bo) - return -ENOMEM; + if (IS_ERR(exec->tile_bo)) + return PTR_ERR(exec->tile_bo); list_add_tail(&tile_bo->unref_head, &exec->unref_list); /* tile alloc address. */ |
