/* * vsp1_drm.c -- R-Car VSP1 DRM API * * Copyright (C) 2015 Renesas Electronics Corporation * * Contact: Laurent Pinchart (laurent.pinchart@ideasonboard.com) * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. */ #include #include #include #include #include #include "vsp1.h" #include "vsp1_bru.h" #include "vsp1_dl.h" #include "vsp1_drm.h" #include "vsp1_lif.h" #include "vsp1_pipe.h" #include "vsp1_rwpf.h" /* ----------------------------------------------------------------------------- * Runtime Handling */ static void vsp1_drm_pipeline_frame_end(struct vsp1_pipeline *pipe) { unsigned long flags; spin_lock_irqsave(&pipe->irqlock, flags); if (pipe->num_inputs) vsp1_pipeline_run(pipe); spin_unlock_irqrestore(&pipe->irqlock, flags); } /* ----------------------------------------------------------------------------- * DU Driver API */ int vsp1_du_init(struct device *dev) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); if (!vsp1) return -EPROBE_DEFER; return 0; } EXPORT_SYMBOL_GPL(vsp1_du_init); /** * vsp1_du_setup_lif - Setup the output part of the VSP pipeline * @dev: the VSP device * @width: output frame width in pixels * @height: output frame height in pixels * * Configure the output part of VSP DRM pipeline for the given frame @width and * @height. This sets up formats on the BRU source pad, the WPF0 sink and source * pads, and the LIF sink pad. * * As the media bus code on the BRU source pad is conditioned by the * configuration of the BRU sink 0 pad, we also set up the formats on all BRU * sinks, even if the configuration will be overwritten later by * vsp1_du_setup_rpf(). This ensures that the BRU configuration is set to a well * defined state. * * Return 0 on success or a negative error code on failure. */ int vsp1_du_setup_lif(struct device *dev, unsigned int width, unsigned int height) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); struct vsp1_pipeline *pipe = &vsp1->drm->pipe; struct vsp1_bru *bru = vsp1->bru; struct v4l2_subdev_format format; unsigned int i; int ret; dev_dbg(vsp1->dev, "%s: configuring LIF with format %ux%u\n", __func__, width, height); if (width == 0 || height == 0) { /* Zero width or height means the CRTC is being disabled, stop * the pipeline and turn the light off. */ ret = vsp1_pipeline_stop(pipe); if (ret == -ETIMEDOUT) dev_err(vsp1->dev, "DRM pipeline stop timeout\n"); media_entity_pipeline_stop(&pipe->output->entity.subdev.entity); for (i = 0; i < bru->entity.source_pad; ++i) { bru->inputs[i].rpf = NULL; pipe->inputs[i] = NULL; } pipe->num_inputs = 0; vsp1_device_put(vsp1); dev_dbg(vsp1->dev, "%s: pipeline disabled\n", __func__); return 0; } vsp1_dl_reset(vsp1->drm->dl); /* Configure the format at the BRU sinks and propagate it through the * pipeline. */ memset(&format, 0, sizeof(format)); format.which = V4L2_SUBDEV_FORMAT_ACTIVE; for (i = 0; i < bru->entity.source_pad; ++i) { format.pad = i; format.format.width = width; format.format.height = height; format.format.code = MEDIA_BUS_FMT_ARGB8888_1X32; format.format.field = V4L2_FIELD_NONE; ret = v4l2_subdev_call(&bru->entity.subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on BRU pad %u\n", __func__, format.format.width, format.format.height, format.format.code, i); } format.pad = bru->entity.source_pad; format.format.width = width; format.format.height = height; format.format.code = MEDIA_BUS_FMT_ARGB8888_1X32; format.format.field = V4L2_FIELD_NONE; ret = v4l2_subdev_call(&bru->entity.subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on BRU pad %u\n", __func__, format.format.width, format.format.height, format.format.code, i); format.pad = RWPF_PAD_SINK; ret = v4l2_subdev_call(&vsp1->wpf[0]->entity.subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on WPF0 sink\n", __func__, format.format.width, format.format.height, format.format.code); format.pad = RWPF_PAD_SOURCE; ret = v4l2_subdev_call(&vsp1->wpf[0]->entity.subdev, pad, get_fmt, NULL, &format); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: got format %ux%u (%x) on WPF0 source\n", __func__, format.format.width, format.format.height, format.format.code); format.pad = LIF_PAD_SINK; ret = v4l2_subdev_call(&vsp1->lif->entity.subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on LIF sink\n", __func__, format.format.width, format.format.height, format.format.code); /* Verify that the format at the output of the pipeline matches the * requested frame size and media bus code. */ if (format.format.width != width || format.format.height != height || format.format.code != MEDIA_BUS_FMT_ARGB8888_1X32) { dev_dbg(vsp1->dev, "%s: format mismatch\n", __func__); return -EPIPE; } /* Mark the pipeline as streaming and enable the VSP1. This will store * the pipeline pointer in all entities, which the s_stream handlers * will need. We don't start the entities themselves right at this point * as there's no plane configured yet, so we can't start processing * buffers. */ ret = vsp1_device_get(vsp1); if (ret < 0) return ret; ret = media_entity_pipeline_start(&pipe->output->entity.subdev.entity, &pipe->pipe); if (ret < 0) { dev_dbg(vsp1->dev, "%s: pipeline start failed\n", __func__); vsp1_device_put(vsp1); return ret; } dev_dbg(vsp1->dev, "%s: pipeline enabled\n", __func__); return 0; } EXPORT_SYMBOL_GPL(vsp1_du_setup_lif); /** * vsp1_du_atomic_begin - Prepare for an atomic update * @dev: the VSP device */ void vsp1_du_atomic_begin(struct device *dev) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); struct vsp1_pipeline *pipe = &vsp1->drm->pipe; unsigned long flags; spin_lock_irqsave(&pipe->irqlock, flags); vsp1->drm->num_inputs = pipe->num_inputs; spin_unlock_irqrestore(&pipe->irqlock, flags); /* Prepare the display list. */ vsp1_dl_begin(vsp1->drm->dl); } EXPORT_SYMBOL_GPL(vsp1_du_atomic_begin); /** * vsp1_du_atomic_update - Setup one RPF input of the VSP pipeline * @dev: the VSP device * @rpf_index: index of the RPF to setup (0-based) * @pixelformat: V4L2 pixel format for the RPF memory input * @pitch: number of bytes per line in the image stored in memory * @mem: DMA addresses of the memory buffers (one per plane) * @src: the source crop rectangle for the RPF * @dst: the destination compose rectangle for the BRU input * * Configure the VSP to perform composition of the image referenced by @mem * through RPF @rpf_index, using the @src crop rectangle and the @dst * composition rectangle. The Z-order is fixed with RPF 0 at the bottom. * * Image format as stored in memory is expressed as a V4L2 @pixelformat value. * As a special case, setting the pixel format to 0 will disable the RPF. The * @pitch, @mem, @src and @dst parameters are ignored in that case. Calling the * function on a disabled RPF is allowed. * * The memory pitch is configurable to allow for padding at end of lines, or * simple for images that extend beyond the crop rectangle boundaries. The * @pitch value is expressed in bytes and applies to all planes for multiplanar * formats. * * The source memory buffer is referenced by the DMA address of its planes in * the @mem array. Up to two planes are supported. The second plane DMA address * is ignored for formats using a single plane. * * This function isn't reentrant, the caller needs to serialize calls. * * TODO: Implement Z-order control by decoupling the RPF index from the BRU * input index. * * Return 0 on success or a negative error code on failure. */ int vsp1_du_atomic_update(struct device *dev, unsigned int rpf_index, u32 pixelformat, unsigned int pitch, dma_addr_t mem[2], const struct v4l2_rect *src, const struct v4l2_rect *dst) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); struct vsp1_pipeline *pipe = &vsp1->drm->pipe; const struct vsp1_format_info *fmtinfo; struct v4l2_subdev_selection sel; struct v4l2_subdev_format format; struct vsp1_rwpf_memory memory; struct vsp1_rwpf *rpf; unsigned long flags; int ret; if (rpf_index >= vsp1->info->rpf_count) return -EINVAL; rpf = vsp1->rpf[rpf_index]; if (pixelformat == 0) { dev_dbg(vsp1->dev, "%s: RPF%u: disable requested\n", __func__, rpf_index); spin_lock_irqsave(&pipe->irqlock, flags); if (pipe->inputs[rpf_index]) { /* Remove the RPF from the pipeline if it was previously * enabled. */ vsp1->bru->inputs[rpf_index].rpf = NULL; pipe->inputs[rpf_index] = NULL; pipe->num_inputs--; } spin_unlock_irqrestore(&pipe->irqlock, flags); return 0; } dev_dbg(vsp1->dev, "%s: RPF%u: (%u,%u)/%ux%u -> (%u,%u)/%ux%u (%08x), pitch %u dma { %pad, %pad }\n", __func__, rpf_index, src->left, src->top, src->width, src->height, dst->left, dst->top, dst->width, dst->height, pixelformat, pitch, &mem[0], &mem[1]); /* Set the stride at the RPF input. */ fmtinfo = vsp1_get_format_info(pixelformat); if (!fmtinfo) { dev_dbg(vsp1->dev, "Unsupport pixel format %08x for RPF\n", pixelformat); return -EINVAL; } rpf->fmtinfo = fmtinfo; rpf->format.num_planes = fmtinfo->planes; rpf->format.plane_fmt[0].bytesperline = pitch; rpf->format.plane_fmt[1].bytesperline = pitch; /* Configure the format on the RPF sink pad and propagate it up to the * BRU sink pad. */ memset(&format, 0, sizeof(format)); format.which = V4L2_SUBDEV_FORMAT_ACTIVE; format.pad = RWPF_PAD_SINK; format.format.width = src->width + src->left; format.format.height = src->height + src->top; format.format.code = fmtinfo->mbus; format.format.field = V4L2_FIELD_NONE; ret = v4l2_subdev_call(&rpf->entity.subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on RPF%u sink\n", __func__, format.format.width, format.format.height, format.format.code, rpf->entity.index); memset(&sel, 0, sizeof(sel)); sel.which = V4L2_SUBDEV_FORMAT_ACTIVE; sel.pad = RWPF_PAD_SINK; sel.target = V4L2_SEL_TGT_CROP; sel.r = *src; ret = v4l2_subdev_call(&rpf->entity.subdev, pad, set_selection, NULL, &sel); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: set selection (%u,%u)/%ux%u on RPF%u sink\n", __func__, sel.r.left, sel.r.top, sel.r.width, sel.r.height, rpf->entity.index); /* RPF source, hardcode the format to ARGB8888 to turn on format * conversion if needed. */ format.pad = RWPF_PAD_SOURCE; ret = v4l2_subdev_call(&rpf->entity.subdev, pad, get_fmt, NULL, &format); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: got format %ux%u (%x) on RPF%u source\n", __func__, format.format.width, format.format.height, format.format.code, rpf->entity.index); format.format.code = MEDIA_BUS_FMT_ARGB8888_1X32; ret = v4l2_subdev_call(&rpf->entity.subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; /* BRU sink, propagate the format from the RPF source. */ format.pad = rpf->entity.index; ret = v4l2_subdev_call(&vsp1->bru->entity.subdev, pad, set_fmt, NULL, &format); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: set format %ux%u (%x) on BRU pad %u\n", __func__, format.format.width, format.format.height, format.format.code, format.pad); sel.pad = rpf->entity.index; sel.target = V4L2_SEL_TGT_COMPOSE; sel.r = *dst; ret = v4l2_subdev_call(&vsp1->bru->entity.subdev, pad, set_selection, NULL, &sel); if (ret < 0) return ret; dev_dbg(vsp1->dev, "%s: set selection (%u,%u)/%ux%u on BRU pad %u\n", __func__, sel.r.left, sel.r.top, sel.r.width, sel.r.height, sel.pad); /* Store the compose rectangle coordinates in the RPF. */ rpf->location.left = dst->left; rpf->location.top = dst->top; /* Set the memory buffer address. */ memory.num_planes = fmtinfo->planes; memory.addr[0] = mem[0]; memory.addr[1] = mem[1]; rpf->ops->set_memory(rpf, &memory); spin_lock_irqsave(&pipe->irqlock, flags); /* If the RPF was previously stopped set the BRU input to the RPF and * store the RPF in the pipeline inputs array. */ if (!pipe->inputs[rpf->entity.index]) { vsp1->bru->inputs[rpf_index].rpf = rpf; pipe->inputs[rpf->entity.index] = rpf; pipe->num_inputs++; } spin_unlock_irqrestore(&pipe->irqlock, flags); return 0; } EXPORT_SYMBOL_GPL(vsp1_du_atomic_update); /** * vsp1_du_atomic_flush - Commit an atomic update * @dev: the VSP device */ void vsp1_du_atomic_flush(struct device *dev) { struct vsp1_device *vsp1 = dev_get_drvdata(dev); struct vsp1_pipeline *pipe = &vsp1->drm->pipe; struct vsp1_entity *entity; unsigned long flags; bool stop = false; int ret; list_for_each_entry(entity, &pipe->entities, list_pipe) { /* Disconnect unused RPFs from the pipeline. */ if (entity->type == VSP1_ENTITY_RPF) { struct vsp1_rwpf *rpf = to_rwpf(&entity->subdev); if (!pipe->inputs[rpf->entity.index]) { vsp1_mod_write(entity, entity->route->reg, VI6_DPR_NODE_UNUSED); continue; } } vsp1_entity_route_setup(entity); ret = v4l2_subdev_call(&entity->subdev, video, s_stream, 1); if (ret < 0) { dev_err(vsp1->dev, "DRM pipeline start failure on entity %s\n", entity->subdev.name); return; } } vsp1_dl_commit(vsp1->drm->dl); spin_lock_irqsave(&pipe->irqlock, flags); /* Start or stop the pipeline if needed. */ if (!vsp1->drm->num_inputs && pipe->num_inputs) { vsp1_write(vsp1, VI6_DISP_IRQ_STA, 0); vsp1_write(vsp1, VI6_DISP_IRQ_ENB, VI6_DISP_IRQ_ENB_DSTE); vsp1_pipeline_run(pipe); } else if (vsp1->drm->num_inputs && !pipe->num_inputs) { stop = true; } spin_unlock_irqrestore(&pipe->irqlock, flags); if (stop) { vsp1_write(vsp1, VI6_DISP_IRQ_ENB, 0); vsp1_pipeline_stop(pipe); } } EXPORT_SYMBOL_GPL(vsp1_du_atomic_flush); /* ----------------------------------------------------------------------------- * Initialization */ int vsp1_drm_create_links(struct vsp1_device *vsp1) { const u32 flags = MEDIA_LNK_FL_ENABLED | MEDIA_LNK_FL_IMMUTABLE; unsigned int i; int ret; /* VSPD instances require a BRU to perform composition and a LIF to * output to the DU. */ if (!vsp1->bru || !vsp1->lif) return -ENXIO; for (i = 0; i < vsp1->info->rpf_count; ++i) { struct vsp1_rwpf *rpf = vsp1->rpf[i]; ret = media_create_pad_link(&rpf->entity.subdev.entity, RWPF_PAD_SOURCE, &vsp1->bru->entity.subdev.entity, i, flags); if (ret < 0) return ret; rpf->entity.sink = &vsp1->bru->entity.subdev.entity; rpf->entity.sink_pad = i; } ret = media_create_pad_link(&vsp1->bru->entity.subdev.entity, vsp1->bru->entity.source_pad, &vsp1->wpf[0]->entity.subdev.entity, RWPF_PAD_SINK, flags); if (ret < 0) return ret; vsp1->bru->entity.sink = &vsp1->wpf[0]->entity.subdev.entity; vsp1->bru->entity.sink_pad = RWPF_PAD_SINK; ret = media_create_pad_link(&vsp1->wpf[0]->entity.subdev.entity, RWPF_PAD_SOURCE, &vsp1->lif->entity.subdev.entity, LIF_PAD_SINK, flags); if (ret < 0) return ret; return 0; } int vsp1_drm_init(struct vsp1_device *vsp1) { struct vsp1_pipeline *pipe; unsigned int i; vsp1->drm = devm_kzalloc(vsp1->dev, sizeof(*vsp1->drm), GFP_KERNEL); if (!vsp1->drm) return -ENOMEM; vsp1->drm->dl = vsp1_dl_create(vsp1); if (!vsp1->drm->dl) return -ENOMEM; pipe = &vsp1->drm->pipe; vsp1_pipeline_init(pipe); pipe->frame_end = vsp1_drm_pipeline_frame_end; /* The DRM pipeline is static, add entities manually. */ for (i = 0; i < vsp1->info->rpf_count; ++i) { struct vsp1_rwpf *input = vsp1->rpf[i]; list_add_tail(&input->entity.list_pipe, &pipe->entities); } list_add_tail(&vsp1->bru->entity.list_pipe, &pipe->entities); list_add_tail(&vsp1->wpf[0]->entity.list_pipe, &pipe->entities); list_add_tail(&vsp1->lif->entity.list_pipe, &pipe->entities); pipe->bru = &vsp1->bru->entity; pipe->lif = &vsp1->lif->entity; pipe->output = vsp1->wpf[0]; pipe->dl = vsp1->drm->dl; return 0; } void vsp1_drm_cleanup(struct vsp1_device *vsp1) { vsp1_dl_destroy(vsp1->drm->dl); }